By on September 2, 2011

The Environmental Protection Agency’s fuel economy testing system is notoriously weak, relying on self-reporting for the vast majority of vehicles, and exhibiting vulnerabilities to “gaming.” But rather than attacking each others’ EPA numbers, automakers seem to have agreed that it’s best if everyone does their best to juice their own numbers and allows the imperfect system to limp on. But over at Automotive News [sub], we’re hearing what could be the first shots fired in a new war over EPA ratings, as Product Editor Rick Kranz reveals that an OEM is starting to complain about another OEM’s fuel economy ratings. He writes:

An executive of one U.S. automaker suggests there might be some sleight of hand going on and that the EPA is not catching the offenders.

The issue: There’s a noticeable difference between the mpg number posted on some cars’ window sticker and an analysis of the data submitted by automakers to the EPA.


Kranz continues:

The executive raised a red flag earlier this year. He told me his company was unable to replicate the city, highway and overall fuel economy numbers achieved by some automakers for their 2011 car models.

He didn’t name the automakers or the car models in question. Neither would he give the percentage differences between the mpg numbers posted on new-car window stickers and an analysis of the data taken from dynamometer readings his company purchased for certain competing models.

But he said consumers are being misled. The mpg numbers on some window stickers or in advertising are being misrepresented, he said.

Here’s the thing: if an executive is complaining about another OEM gaming the EPA test or somehow fudging its results, this executive must be extremely angry or frustrated. After all, a weak EPA testing regime benefits all automakers at the expense of customers. And if someone is willing to blow down the EPA’s house of cards, there’s no knowing where the fallout could end. There are basically three possibilities:
1) The accusing executive has the wrong end of the stick, and is just lashing out without cause.
2) The accusing executive is on to something and an automaker is fudging its EPA numbers.
3) The accusing executive is on to something, and he’s just scratching the surface of a problem infecting a large part of the industry.
As fuel economy becomes a bigger factor in car-buying decisions, the EPA needs to recognize that there is more riding on its weak, “faith-based” fuel economy testing regime than ever. It should not only investigate this allegation, but it should perform supplemental targeted verification tests on vehicles with “suspiciously high” fuel economy ratings. Consumers need to trust their window stickers, and if there are rumors of gamesmanship around the production of those numbers, competitive pressure will spread deceptive practices around the industry. This needs to be nipped on the bud.
So, in hopes of helping the EPA get a handle on this situation, I ask the B&B to share their thoughts about what automakers might be fudging their numbers. What vehicles would you spot-test to see if they can achieve their window sticker numbers?
Get the latest TTAC e-Newsletter!

95 Comments on “Ask The Best And Brightest: Which Automaker May Be Fudging Their EPA Numbers?...”

  • avatar

    I have no knowledge of who’s fudging what numbers, but my personal observations from various magazines and sites like Edmumds that do keep track of mileage is thus;

    Tested Hyundai, VW and to a lesser extent Ford vehicles tend to be very close, right on or a little better than the EPA estimates.

    GM,do not. The biggest culprits seem to be the versions of the Equinox and Traverse – Nobody has been able to replicate the Equinox’s 4cyl highway number.

    • 0 avatar
      Secret Hi5

      Edmunds’ Insideline just released the fuel econ numbers for their Long Term test fleet, and compared to official MPGs, the Kia Optima, Fusion Hybrid, Mustang, and Nissan Juke are bad.

      Most of their vehicles fall short of EPA estimates, but in previous months, the Buick Regal *exceeded* the official numbers! Also from past reports, the Terrain and Traverse were gas hogs.

    • 0 avatar

      The Hyundai Elantra hasn’t been so great at matching its best-in-class EPA ratings – Car and Driver got lower mileage than the Focus in a comparo, and Consumer Reports got 39mpg on the highway (the Focus got 43, and the new Civic got an astounding 47).

    • 0 avatar
      Sgt Beavis

      It should be mentioned that the EPA mileage results are based on cars using 100% gasoline. Not the E10 crap that we have to use in our tanks. That is done in accordance with EPA rules.

      E10 will lower your mileage performance. The question should be whether or not it is enough for these kinds of discrepancies.

      BTW, I own a 2010 Equinox. 30mpg is the best I’ve ever managed on the highway. 28mpg is the norm. However I am otherwise very happy with the car.

      • 0 avatar

        E10 will lower your mileage performance. The question should be whether or not it is enough for these kinds of discrepancies.

        The difference is about 3%. For a car getting 25 mpg, this would equal a whopping 0.75 mpg.

      • 0 avatar

        The difference can be dramatic. In my own late model EFI vehciles it is around 10%.

        Years go I read a study checking the effects of E10 on MPG and the worst of the 8 cars was a Ford Fusion with a whopping 18% drop in MPG between real gas and E10. The Impala which was a FFV actually increased it’s MPG 2 or 3% on E10.

        After the OR gov’t mandated it one of the legislators, who voted for the bill in the first place, noted her Prius lost between 4 and 5 MPG. So she introduced bills to reverse. She finally got an exemption for premium but it can only be dispensed into 25 year old vehicles so it pretty much isn’t available. For those that don’t know self serve gas is not legal in OR, so the station could be fined should their employees dispense it into a car that isn’t old enough.

      • 0 avatar

        I read a study checking the effects of E10 on MPG and the worst of the 8 cars was a Ford Fusion with a whopping 18% drop in MPG between real gas and E10.

        I’d like to see this alleged study, as that claim makes no sense whatsoever.

        The energy content of ethanol by volume is about 30% less than the equivalent amount of gasoline. E10 is 10% ethanol, so the resulting loss of energy is 3%. (0.3 X 0.1 = .03) All things being equal, fuel consumption by volume = energy content.

        For a car to get 18% less mileage on E10 would be equivalent to saying that ethanol contained a negative amount of energy, which of course is impossible.

        As the EPA notes, “Most conventional vehicles using E10 (10 percent ethanol) will experience a 3 to 4 percent reduction in fuel economy.” That is a logical figure, given the energy content of the fuel.

    • 0 avatar

      Hyundai (and by relation, that Optima) vehicles are definitely not that close to their rated EPA numbers, and from what I understand the Sonata Hybrid is particularly sad compared to the rest of the hybrids in the class.

    • 0 avatar

      Ethanol absorbs water (which would corrode petrol pipelines), and thus is “splashed” into the tanker at the terminal (which absorbs more water), and the percentage mixed in gasoline (I suspect) would be higher than claimed, especially since it’s a subsidized, cheaper alternative to gasoline.
      I also suspect that the benefits of direct injection are diminished (unless extensive tests of ethanol blends were performed) by the altered combustion chamber temperature caused by ethanol.

      So I’m a suspicious mug by nature… but the nearest gas station that claims to offer “pure” gasoline is about 40 miles northeast of here, so it’s out of range of any (same weather, same driving conditions, etc.) reasonable, longer-term comparison for me.

      IOW, someone should actually try to perform this comparison to see what’s up. Whaddaya think TTAC?

      • 0 avatar

        There may be just a 3% drop in energy content E10 vs straight gasoline, but are the ECUs programmed for the most efficient combustion? How about the combustion chamber geometry? Summing up joules of energy per kg doesn’t account for that.
        On the other hand, you have lead foots who are disappointed with their mileage and also would be if they were using pure gasoline.
        And let’s not forget that due to tolerancing stackups, engines will have some variability in efficiency.
        Maybe TTAC could get Jack Baruth and David Holzman to test the same car with and without ethanol on the same course on the same day. Average the results for the real world driver. Gets complicated and expensive fast.

      • 0 avatar

        There may be just a 3% drop in energy content E10 vs straight gasoline, but are the ECUs programmed for the most efficient combustion?

        Yes, they are.

        “TOP TIER Detergent Gasoline is the premier standard for gasoline performance. Six of the world’s top automakers, BMW, General Motors, Honda, Toyota, Volkswagen and Audi recognize that the current EPA minimum detergent requirements do not go far enough to ensure optimal engine performance.”

        And, yep, you guessed it, Top Tier Gas is E10.

        A modern car is designed to run on E10. The ethanol has detergent value, so at those levels, it helps a modern car to run better.

      • 0 avatar

        Ok so I was a little off in my recollection of the exact results but here is the link to the 36 page report. here is a condensed version but since it is from ACE I’m sure you’ll say it is bunk even though they are using the results from the first link. Oops forgot this link the first time

        As you can see using the BTU content to calculate the difference in fuel economy like the EPA does for it’s E85 ratings of FFV is worthless. Ethanol is a different fuel than gas and as such it needs different timing curve which plays a large factor in engine efficiency.

        Personally I have done my own less scientific testing. I’ve driven my standard late model FI cars on trips in and out of area that have pure gas and mandated or oil company enriching E10. My results were that I loose about 10% in MPG going to E10.

        I have also done limited testing of the higher blends in the FFV we own. The first time that gas approached $4/gal I decided that a 15MPG SUV was a little expensive for the amount of driving my wife did with it. So I went looking for a beater sedan for the bulk of her driving. A very small local dealer had a Taurus FFV that she wanted to “wholesale to the public”. She got it in trade and since she still mainly sold SUVs and the fact that it was rough around the edges she didn’t want it for her lot. (She only stocks 5-6 vehicles normally). It was very dirty had a broken tail light, tons of scratches on the trunk lid and needed front brakes bad. After a half a day cleaning the interior, buffing the trunk, replacing the tail light and doing the brakes I had a car for a little less than half it’s value.

        I didn’t buy it because it was a FFV I bought it because I knew I could drive it for a couple of years for free and it would save money on the fuel bill.

        E85 wasn’t available in my immediate area but is available not that far away. So when I took it on a trip through areas that have E85 I experimented. I found that like the tests I linked to my MPG with somewhere in the E40-E50 range the MPG was better than the E10. The first time I did it was sort of accidental. I had about a half a tank of E85 and topped it off with gas since I was going to be away from any large cities for the next leg of the trip so I didn’t want to be stuck over paying at the sole station out in the middle of nowhere. Much to my surprise on the next fill up I noticed it was the best MPG I had achieved. I dismissed it as a bit of a fluke. Then when I got home I did some investigating and found that first link. Since then I’ve done “splash blending” testing multiple times with that vehicle. Every time the result is the same with E40~E50 in the tank it delivers 12~13% better MPG than on straight gas. It also makes more power which may account for some of that. The boost in low end torque allows it to climb grades in OD that it can’t with gas. My wife even noticed the difference when I brought it back with summer blend E85 (70% ethanol). She drove it and when she got home she said “what did you do to my car”. I sheepishly said “nothing” thinking she found a dent or something. She said so you didn’t do any modifications (my cars are never kept stock). I said no. She so you didn’t tune it up or anything because it has a lot more power. I replied only the power of corn whiskey.

        She’s a convert now and we are taking it on our trip to the coast later today and it will be running E40~E50 as much as I can since we will pass a couple of stations that carry E85 along the way. I end up saving about 10% on price, get better MPG and more power too.

      • 0 avatar

        Ethanol absorbs water (which would corrode petrol pipelines), and thus is “splashed” into the tanker at the terminal (which absorbs more water), and the percentage mixed in gasoline (I suspect) would be higher than claimed, especially since it’s a subsidized, cheaper alternative to gasoline.

        For E10, that amounts to about 0.5% or less, depending upon the temperature. (As the temperature drops, the absorption rate falls with it.) It’s not much of a factor.

        As you can see using the BTU content to calculate the difference in fuel economy like the EPA does for it’s E85 ratings of FFV is worthless.

        You should have read the study more closely. The MPG fuel economies have a downward trendline to match the increase in ethanol intake. The “calculated” result is based upon the energy content, and the actual results generally follow the same direction.

        In any case, any study will have individual anomalies in it. If you notice the individual test results for each car, the three Tier 2 gas tests for the Impala had differences between the highest and the lowest result of about 6%, while the Fusion had differences on Tier 2 gas of about 3%, despite using the same fuel. In comparison, the Camry’s actual results tracked the calculated result almost perfectly. Although the lab will attempt to completely control all conditions, the results can still be flawed. The outliers are just that – outliers — and exceptions don’t disprove the rule.

      • 0 avatar

        PCH I have read the study closely and in all cases the actual observed mileage does not correspond to the MPG calculated by using BTUs. My own extensive testing of different blends in my FFV and E10 vs gas on my non FFV vehicles supports the facts of the test. As stated E40-E50 consistently gives repeatable increases over E10 while using the BTU method says that the MPG should be lower with the higher amount of ethanol. Also other than some Dodges I have never seen anyone report a drop in MPG that corresponds to the calculated difference in Gas vs E85 MPG as listed by the EPA.

        At low cylinder pressures (part throttle) fuel containing ethanol burns slower than gas while at high cylinder pressures (WOT) it burns quicker than gas. Which is why the Impala FFV has different results than the Non FFV version despite the basic engine being the same.

        Call it anecdotal or outliers if you will but that doesn’t change the fact that using BTUs to compare ICE fuels is a flawed idea.

      • 0 avatar

        I have read the study closely and in all cases the actual observed mileage does not correspond to the MPG calculated by using BTUs

        If you had read it closely, then you wouldn’t make that comment.

        Figure 10 compares the actual with the calculated results of the Camry. For E10, E20, E40, E50 and E60, the two are almost exactly the same. In other words, the test results correlated with the projection based upon energy content.

        Figure 11 shows this for the non-FFV Impala. The two are basically the same for E10, E20, E30 and E50, and just slightly different for E40.

        Figures 12 and 13 are for the FFV Impala and Fusion. Both lines deviate from the calculated result, but they also trend downward, showing that higher ethanol content generally corresponds with lower mileage.

        And as I noted about the FFV Impala and Fusion, the two of these had mileage differences during their tests of E0. The difference between the highest and lowest mpg result for the FFV Impala on 100% gasoline was 6%, which suggests that there may have an issue with the test controls and/ or the car. (It helps to read the stuff in the back, you know.)

        My own extensive testing of different blends in my FFV and E10 vs gas

        “Extensive testing”? Are you now claiming to be a trained researcher?

      • 0 avatar

        Top Tier Gas is not required to contain ethanol, but the deposit control testing must be done using E10 to ensure compliance in a worst-case scenario. I can buy ethanol-free premium (91) Top Tier Gas in my area. The 87 is E10.

      • 0 avatar

        Top Tier Gas is not required to contain ethanol

        That is incorrect. Top Tier Gas must have 8-10% ethanol:

        These standards comprise the requirements for TOP TIER Detergent Gasoline… Base Fuel. The base fuel shall conform to ASTM D 4814 and shall contain commercial fuel grade ethanol conforming to ASTM D 4806…The base fuel shall have the following specific properties after the addition of ethanol:

        1. Contain enough denatured ethanol such that the actual ethanol content is no less than 8.0 and no more than 10.0 volume percent.

  • avatar

    “[Kranz] didn’t name the automakers or the car models in question. Neither would he give the percentage differences between the mpg numbers posted on new-car window stickers and an analysis of the data taken from dynamometer readings his company purchased for certain competing models.”

    It would help us a lot if did give us the names and the specifics. But of course he doesn’t want to be blackballed by the OEMs.

    • 0 avatar

      Unfortunately I agree, but I can’t see how Kranz expects to have any credibility here. He’s not really risking anything important by naming names, other than his ability to feel special. Yes, the anonymous excec might get an earful for talking to a reporter, who then in turn might blackball Kranz.

      But like EN said, this isn’t idle gossip. The public has an interest in accurate MPG ratings, so if there is systematic cheating then that would be, to say the least, newsworthy.

      More likely that the OEM knows exactly what the exec is doing, and is intentionally using Kranz as a mouthpiece while insisting on anonymity, thereby avoiding any risk of fallout.

      So either Kranz knows the story is false and he’s repeating it anyway to fill up column space, or he thinks it’s true and he’s selling out at the expense of the public interest. Not good options.

  • avatar

    I’m not so sure about this self reporting thing. The way I understand it the testing is done by an independent 3rd party that is certified by the EPA to do the combined MPG/emissions testing cycle. Sure it would be possible to set up the engine calibration purely to maximize the MPG and minimize the emissions on that carefully controlled test, that may not come anywhere near reflecting the real world. There is also the possibility that they mfg tweaked the calibration after testing. GM and Honda got busted for doing that back in the 90’s after too many of their vehicles set off flags in state emissions testing.

    GM issued a revised calibration for Cadillacs that switched to an open loop idle that was rich when the AC was on to address rough idle complaints.

    Honda just never sold cars with the calibration that was used for testing. Instead they ran them lean during cruise to increase the real world MPG at the expense of emissions.

  • avatar

    Consumers have accused carmakers of EPA shenanigans since the 1970s, and I thought the revised rules in 2008 were supposed to address this.

    I’d spot-test anything claiming over 40 mpg highway. However, as highway MPGs continue to climb, so does their sensitivity to real-world conditions, and it will become increasingly difficult for consumers to match mfr’s claims. For example, a 5% drop on a ’40 mpg’ car is 2 mpg, while the same 5% amounts to only a 1 mpg drop on a ’20 mpg’ car.

    • 0 avatar

      A 5% drop on a 40 mpg car is not necessarily more than a 5% drop on a 20 mpg car. For instance, the difference between 42mpg and 45mpg is a lot less than the difference between 12mpg and 15mpg.

      • 0 avatar

        Except that you described a 6.7% drop vs a 20% drop in the second sentence, so of course they’re not the same.

        And even though the extra fuel used by the same 5% drop on a 40-mpg car is less than the extra fuel used by a 5% drop in a 20-mpg car, consumers perceive the 2 mpg difference as ‘worse’ than the 1 mpg difference of the guzzler.

        Therefore, consumer perception is an important consideration when people argue about how faithful their cars are to EPA ratings – on this point I think we agree.

      • 0 avatar

        MPG is bass-ackwards. We need L/100km or Gallons/100miles to determine efficiency, which I think is x9419c is trying to get at.

  • avatar

    From my own very small sample space:

    These cars seemed right on :

    2004 Chrysler Sebring – epa hwy 27? … I got 29-30
    2002 VW Eurovan – epa hwy 19 … I got 19 … when not in the shop, which was always
    1996 Subaru Outback – epa hwy 29 … I got 29-32

    My Hondas have been under rated :

    1993 Civic – epa hwy 40 … I got 48-50
    2005 Civic – epa hwy 38 … I get 40-43

    This car is not even close:

    2006 Toyota Matrix – epa hwy 36 … I get 30-33

    All of them operated under roughly the same conditions and driving style.

  • avatar

    “After all, a weak EPA testing regime benefits all automakers at the expense of customers.”

    You assume that all mfrs are cheating on their mileage numbers, more or less equally, in a gentlemen’s agreement to ignore the cheating. Is it not possible that one or more mfrs are reporting mileage honestly, at least currently? And wants to gain an advantage by getting the EPA to officially call out the cheaters, now that MPG is again becoming a major feature for buyers?

    • 0 avatar

      Yes, which is why I included three possible scenarios. But if even one is cheating, there would be immense pressure on others to either cheat or report the cheating. Because of this pressure, the EPA needs to take even whispered accusations very seriously.

  • avatar

    You mean the car-buying public is being misled? Inconceivable!

  • avatar

    I can’t say which automakers would be “gaming” the system or fiddling with the numbers around the edges, but the only number that I have ever really paid attention to is the combined rating.

    For instance when I got my car, the combined rating was 29mpg. I am able to match this pretty much all the time +/- 1 mpg. If I’m driving exceptionally wastefully or exuberantly it’ll go down to 27ish. I only pay attention to this number because I will rarely have a complete tank that is freeway, and I will never have a complete tank that is city.

    Now I have had experience where I was doing to 80 mph for a 300 mile roundtrip, in one day, and managed a nice 36.XX which is 2.XX above the highway rating.

    All that being said, I wouldn’t be able to credibly think of a good way to reduce the reported estimates.

  • avatar

    i’d be surprised if it wasn’t directed at hyundai and kia. all the press cars of theirs i’ve driven in have never come close to advertised MPG, especially the optima.

    to the person who mentioned the 4cylinder equinox/terrain, i’d also agree with that. that thing got terrible mileage and had a terrible drivetrain.

    • 0 avatar

      Actually, I’ve heard of Hyundai’s press cars far exceeding EPA numbers – just the opposite of your experience.

      But since Hyundai/Kia are the corporate MPG leaders, they’d make a natural target.

      • 0 avatar

        It comes down to whether the reviewer fell for Hyundai’s trip computer reporting or not. The test results using gas receipts and long division are never impressive. Just goes to show why the manufacturers play games: Most people are easy to mislead.

  • avatar

    It’s the Ford against Hyundai.

  • avatar

    Avis has been getting a lot of Chevy Cruzes recently. I’ve driven several, including one 200+ mile all-highway jaunt. I got 26 mpg, using the Cruze’s cruise (stupid name!). It was a decent car otherwise, definitely more in its element no the highway (around town, the darn thing feels utterly gutless unless the gearbox kicks down 2-3 gears and shoots the engine over 3500 rpm, at which point you get far too much acceleration). Horribly calibrated vehicle. I thought they might have paid more attention to EPA numbers when engineering the driving experience than actual user-friendliness, especially the highway rating.

    On the return trip, exactly the same number of miles, with several sitting in thickening traffic in a 2011 Charger, I nabbed just over 30 mpg.

    I’ve always felt GM’s EPA numbers were too high. I had one trip from here to a DC airport (230 mi) in a ’10 CTS4 where I AVERAGED just over 18 mpg, all highway, using cruise, and A/C wasn’t on the whole time. The return trip in a Camry, driven somewhat more aggressively, hit 33 mpg. I know this is anecdotal, but I’ve driven tens of thousands of miles in all different types of newer rental vehicles, so it wasn’t just one or two 5-mile trips in a dealer vehicle.

    GM, Mazda, and Suzuki have never lived up to their claims in my experiences.

    Honda, Toyota, Ford (except for the new Explorer V6 – gas hog), and Mitsubishi (!) all do considerably better.

    Everyone else pretty much nabs it or barely misses EPA figures, with a few exceptions (the Hyundai Azeras I’ve driven drank fuel like it was going out of style).

    • 0 avatar

      Wait, 26mpg highway in a Chevy Cruze which has a 35-42mpg highway rating? That number seems impossibly low. Were you going 120mph? Did you do the whole way in 2nd gear? Was there a leak in the gas tank?

  • avatar

    I could never get the claimed mileage from my Mazda3 either in town or on the highway. The short gearing didn’t help on the highway where any additional MPH over 65 was punished with significant drops in fuel economy.

    My current 335 and previous 3 BMWs all exceeded their claimed mileage when driven normally. Even slightly tuned, the 335 gets 30 MPG+ highway at 75 MPH.

    • 0 avatar

      I had the same experience with both manufacturers.

      With my ’02 Protege I couldn’t touch the highway estimates unless I drove the speed limit or under. My ’98 540 is a couple of MPGs better than EPA at 75-80mph, and goes up further the closer you are to the speed limit.

      • 0 avatar

        You can’t expect to get EPA highway while going 80mph, it’s just not designed to reflect that. I was averaging 38 mpg in my 2.0L Mazda 3. It dropped 1 mpg when I got new tires. This was before I was really practicing pulse/glide when the interstate is clear. I don’t record numbers anymore because I’ve been doing some engine-off coasting at certain points in my commute, and the odometer/trip doesn’t record mileage when the engine is off.

      • 0 avatar

        Demetri I’ve done better than or met the “old” EPA hwy at 80 MPH or even higher, in many of the Ford products I’ve had, when using real gas.

      • 0 avatar


        How do you do engine-off coasting in a Mazda3 without the car freaking out? I tried it once on a long offramp. I coasted roughly 3/4 of a mile then I turned the car back on. Power steering was disabled and a couple of warning lights came on. Luckily everything reset itself with a key off/on cycle.

      • 0 avatar

        Scout, you may get it, but don’t expect it, because the EPA test doesn’t reflect speeds that high. I think the test only goes up to 60mph. Gearing and drag coefficient play a big role at those high speeds.

        Mazder, the power steering won’t always come back on, in fact, many times it doesn’t, and in that case the power steering light may pop up. I’ve become accustomed to driving without it. It will turn back on suddenly during uphill elevation changes and hard steering angles, which is important to be aware of if you do any engine off driving. The only other issue I’ve had is that one time after re-starting the engine while driving, the fuel gauge reset to empty. Over the course of about 40 miles, it had re-calibrated itself back to normal.

  • avatar

    Speaking from ownership I can say there is little difference between a Versa and Sentra SER – Sentra with larger 4. The Versa was/is fudged.

  • avatar

    There’s lots of reports of Ford Ecoboost V6’s in the trucks not getting anywhere near their EPA labels.

    However, from my own experience, GM is a (recent) cheat on this. I used to have a 98 Chevy Blazer ZR2 5-speed that rates at 19mpg highway. Around town I averaged just under 18, highway 20, so it did great. My 2004 Mercury Mountaineer AWD V6 also beats the EPA estimate on the highway…I regularly get about 19.5 on a highway trip we frequently take from our doorstep to a vacation home. However, our 2010 Malibu has never come close to it’s 33mpg highway rating. Best I’ve ever done on the same highway trip as the aforementioned Mountaineer is 29 mpg, and at 22k miles, it’s plenty broken in. It’s a shame because the mileage is one of the reasons we bought it, and I’ve rented Impala’s before and gotten nearly 33mpg on a similar trip and was quite impressed with it.

    • 0 avatar

      When I can get real gas for my 03 Mountaineer V8 3.73 gears I can get 20 MPG (barely) at 75-80 MPH with E10 it’s 18. Do you live where you get E10 or real gas.

    • 0 avatar

      On the link posted below 2011 F-150 V6s (they don’t break down between EcoBoost and NA V6s unfortunately) the average combined fuel economy for 2011 models seems to sit between 17 and 18mpg. The combined EPA fuel economy figure for a 4×4 EcoBoost truck is 17, and for a 4×2 it’s 18, so this seems to make sense (and the EcoBoost trucks outsell the NA V6 by a wide margin, so I’m guessing most of these in the survey are EcoBoost models).

      That being said, the fuel economy on the EcoBoost trucks does drop when towing, but the fuel economy of any truck drops when towing. The fuel economy testing is also done with the most favorable available axle ratio. Someone who opts for a 3.73 can’t expect to get the extra towing power and acceleration and still get the same fuel economy as someone with a 3.31.

  • avatar

    Mazda is innocent in regards to the RX-8: EPA is 16/22 and I get that EXACTLY.

    With the 40 mpg bogey so prevalent now, I would be shocked if automakers didn’t do something underhanded to achieve that rating.

    • 0 avatar

      I have 10 yrs of data on my Protege5. Its numbers have been spot-on for the old EPA estimates, and far exceed the new ones. Granted, I don’t drive crazy like other people do, but I believe that is accounted for in the difference between the new & old EPA figures. (The old method predicts performance for drivers like me, and the new method predicts performance for other drivers.)

      One very important point is that I believe with all the new optimized equipment, cars’ efficiencies are becoming increasingly sensitive to how you drive them. We all know if you don’t drive a hybrid correctly, you will never even sniff the mpg numbers. I think the same is becoming true for other cars–try to get good mileage (such as during testing), and you will get it.

  • avatar
    DC Bruce

    Seems like there are two distinct issues here. One is fiddling with the car solely for purposes of maxing out the test, so that the test vehicle is not representative of what’s on sale. That’s cheating and ought to be called out and whopped.

    The other is “gaming the test,” which is setting up the engine so that it maxes out the test but doing other things that use lots of fuel whenever the driver deviates from the test profile in the interest of better performance, etc. To the extent that the test doesn’t replicate the way people drive (and I’m not talking about stoplight drag races), then they’re going to experience different fuel economy from what’s on the sticker.

    In the “suspicious” category, I would put small displacement, turbocharged engines pushing around heavy vehicles. I’ve noticed that both the Acura RDX (2.3 liter 4 powering a 3900 lb. SUV) and new “Ecoboost” Explorer (essentially the same) miss their EPA targets according to those who have tested them (and, some tester reportedly managed to get 11 mpg in the RDX!).

    By contrast, the 2.2 liter turbo 4 in my Saab 9-5 wagon would reliably hit its EPA numbers (31) in highway driving and around town, driven sanely but not timidly.

    • 0 avatar

      Well, this is as anecdotal as as your experience with your Saab, but Mrs. ‘scion consistently beats EPA by 1 or 2 MPG in her ’08 RDX. I’m not sure I could marshall the self control to emulate that myself, but she seems to do it with ease, and continues to wax cheerfully lyrical about power and handling to boot. Her previous ’07 Mini Cooper was right on the EPA nose over 3 years, so I don’t think she’s a closet hypermiler either. Now, granted, if you put your foot in it all the time, the RDX will be as thirsty as you choose, but that surely applies to just about any performance vehicle.

  • avatar

    I doubt that there is much cheating. The EPA doesn’t just rely upon the manufacturers; it also runs its own tests, which should discourage systematic lying.

    What I would expect to be more prevalent would be manufacturers that design their cars in order to game the testing system. But that’s to be expected, of course; any system will encourage gaming. Driving style impacts fuel economy, and since not everybody drives in what could be described as being in an EPA-friendly manner, real-world results can vary quite a bit from driver to driver.

    The companies who would be most likely to be motivated to game the system would be the domestics, due to their unwillingness to pay CAFE fines, combined with their greater dependency on larger cars and trucks for their revenue. The Europeans don’t care about the fines — they sell in low enough volumes and at high enough prices that they can just pay them — while the Asians sell smaller cars that don’t have much problem complying. Still, I doubt that any of the companies is making a widespread practice of it.

  • avatar

    Fusion Hybrid: consumer reports and motor trend both found it to get observed fuel economy way closer to that of other full hybrid midsizers rather than blowing them out of the water as the EPA numbers would suggest.

    Any fullsize truck/SUV with 2 OD gears: No matter what the rating says, every owner I meet of every brand says how they can get the EPA rating only on flat highway driving (perfect conditions). You add any hills or any amount of speed variation and those numbers plummet. This is probably more of a complaint about the EPA highway cycle being pretty easy to game than anything, though.

    Anything that is heavier than their competitors but has way better EPA numbers: Looking at the Chevy Equinox and GMC Terrain here.

  • avatar

    I think this is a non-issue. Too many variables from the test setting of a virtual lab and the real world. E10? Well that hurts MPG. Don’t have perfect tire pressure? That hurts MPG. Change out sticky OEM tires with less grippy ones – may improve MPG. Drive like a grandma, better MPG. Speed racer. Worse MPG. Load in the trunk. Worse MPG. Five-foot tall and weigh 95 pounds? Better MPG.

    Bah – I don’t see anything endemic in the universe.

    • 0 avatar

      I agree. We’ll never find the real killers.

      I forgot about E10 – that alone could be very significant. Not to mention that many people seem to think that driving at 70 mph will cost only a few percentage points in mpg – not so.

      • 0 avatar

        The effect of higher speeds varies wildly depending on the vehicle. If you are driving a minivan with the frontal area and the aerodynamics of a barn going faster will kill you. If you are driving a smaller slippery car with tall gearing the effect is not so pronounced.

        Afterall, if all it took was going slower to get better mileage, then the best mileage would be stopped with the engine running.

      • 0 avatar

        @krhodes1: Fuel economy is a bell curve with respect to speed. The peak of the curve is at some optimal speed, which may be 45 – 65 mph depending on the car.

        A surprising find I had last summer towing with my Sedona was that weight played an enormous role in fuel economy. My total road weight (7 people, camping equipment, coolers, 2-wheel U-Haul) was about 7300 lbs. I averaged about 12 mpg highway driving from PA to CA and back, when normally I can get about 22 mpg without the trailer. All the extra road friction from towing made a huge difference, so highway fuel economy is NOT just about aerodynamics (as I used to think).

      • 0 avatar

        With trailer & without trailer are fundamentally different scenarios and not least because of weight. Trailers will screw up aerodynamics as well as add another set of wheel friction. Weight kills on anything other that flat ground as it takes considerably more energy to raise and you don’t get it back on the downhills because of the need to downshift or brake more to control speed.

        Also, the shape of the mpg curve with speed is not a bell curve (characterized by long tails on either end). MPG rises rapidly at low speeds and then levels off, with the peak typically around 45-60 mph. Since wind resistance is at least quadratic with speed as is engine friction, the drop off after the peak can be very sharp, especially for a weak engine that has to run at high rpm.

  • avatar

    The numbers being reported by this Federal Agency is the most accurate reporting being done by any Federal Agency. Expecting accuracy from any Federal Agency is like expecting virginity in a cat house.

  • avatar

    I’ve owned many cars from many manufacturers and have never had one deliver less highway mpg than the EPA sticker. In fact, most were over, with the 1988 Honda Civic being the overage champ: 45 vs. EPA of 37. Lowest overage experience is the Ford Freestyle, which can eke out perhaps 29 vs. 27 EPA. My BMW M3 gets 30 vs. 27 EPA. 2003 Passat can get 32 mpg which is the EPA estimate. City is another story: Civic was the overage champ at 40, vs. 32 EPA. Ford is under the city estimate, 17 vs. 19. BMW betters the city EPA, 20 vs. 19 mpg. Passat pretty close. Driver matters hugely. I drive pretty conservatively and can get 20-30% better mileage than the other users in our family.

    • 0 avatar

      I’ll second that. My Crossfire got 26 mpg consistently. My Smart got 40-42. My Fiat 500 is getting 37-38.
      The secret: keep the RPM’s down. Drive like you don’t have any brakes.

  • avatar

    Just pump up tire pressures and drive like a grandma and I can see 3-5 mpg increase with a domestic V8. Four cylinders are more sensitive to speed changes as the engine has less reciporating mass, which shows as they are less sensitive to rpms compared to say a V8.

    Car content plays a big role and as a car ages through it’s life cycle tends to gain weight with extra sound deadening and features.

  • avatar

    Our recent trip to St. Louis netted me 35.44 mpg in my 2004 Impala! That’s driving 63-71 mph with the A/C on. On the way home, 33.65 mpg. I’m happy with that.

    I am completing my first full week of my 100-mile-a-day commute, drove the Impala 4 of those days. Right now I have 1/4 tank of gas – actually less gas in the tank than what the gauge reads, but am trying to get four full round trips, although I can stop along the way if I must.

    I’ll report my mileage next week, FWIW. I know as the weather cools off and winter comes, it’ll fall some.

    This commute, once I get to the highway, is against most of the traffic, so it’s a lot of cruising at speed.

    • 0 avatar

      My Malibu Maxx with 3.5 Pushrod V6 and 4 spd autobox I usually got between 30-33 MPG at 75-80 MPH freeway speeds. My 4 cyl & 6 speed autobox G6 gets about 29-31 at the same freeway speeds. I do my fuel calculations by hand, but find that the trip computer is only slightly optomistic, maybe 2 MPG.

      I think the Maxx and G6 are pretty close in weight, and size for sure. I believe I got better fuel mileage with the V6 Chevy because it had more torque and even with the 4 speed tranny it could stay in OD longer because of the amount of torque. The Ecotec does pretty well, but with the two OD ranges on the new 6 speed, I don’t think it has enough grunt under 2200 RPM, at least not like the V6. If I were to keep my speeds slower I know I would get better mileage. It’s all in how you drive and maintain it.

  • avatar

    I can only speak for the Hyundais I regularly drive. I would agree with the Ford vs Hyundai link, you would be hard pressed to get 40mpg on the highway with the Elantra. On a 60 mile drive the best I could do is 37 mpg at about 70 mph, with or without the “eco” button. The Accent, however, got 42 mpg on a 150 mile trip for me – at 75-80mph without the “eco” button. I can easily beat the 26mpg rated average (22 city/35 highway) with my daily driver Sonata – even with the 18″ wheels. I usually average 30 mpg a tank and get over 450 miles per tank. My dad got 41 mpg with his Sonata on his trip to Maine last month, but he drives 65. The 2010 Santa Fe 4 cylinder AWD was rated at 21/27, but the 2011 is rated at 20/25 – lower than the V6!

    • 0 avatar

      I have a coworker who recently purchased a new Elantra for himself and a new Accent for his wife (both automatics) and his experiences bear yours out closely. He finds that the Elantra does mid to high 30’s but the Accent easily tops 40. On a slight tangent, he brought them both by the office to “show off” and they really are very nice cars (potential mpg gaming aside).

    • 0 avatar

      Please – you can’t expect to match EPA numbers while driving at 70-80 mph.

  • avatar

    Why can’t new cars just be filled with ONE GALLON OF GAS and driven regularly until they run out of fuel and STOP?

    In fact, let’s make it more scientific:

    Take 6 different drivers – 2 aggressive, 2 hypermilers and 2 average drivers and compare their distances driven before their car stopped.

    • 0 avatar

      Cars may stop at varying levels of “emptyness”. They don’t necessarily stop when they’re bone-dry.

      And the GM trucks that I’ve owned tend to blow fuel pumps when run dry, as they use fuel as lubricant. :)

    • 0 avatar

      Because of the differences in geometry inside the fuel tanks and of the design of the pickup units. On car may quit with half of the gallon still sloshing around, while another man run until 3/4 of the gallon is consumed.

      Plus you never want to run a fuel-injected car out of gas, as it can destroy the fuel pump.

      I understand what you are trying to say, but I don’t agree that your proposed method would be any more scientific or accurate.

    • 0 avatar

      In fact, let’s make it more scientific

      What you are suggesting would be far less scientific.

    • 0 avatar

      Also, to get statistically meaningful results out of such nebulous testing protocols as “go drive the car regularly” you are going to need way more data than 6 people driving for one gallon’s worth of mileage.

    • 0 avatar

      Back in the day the Shell fuel economy testing was done in a similar fashion. The would disconnect the fuel tank supply line from the fuel pump and run the car until it died. Then hook up their glass jar with 1 gallon, hang it on the door and drive it until it ran out of gas. Which was all find and dandy in the days of carburetors but it just wouldn’t work with modern FI cars.

  • avatar

    GM would be my pick for the biggest culprit.

    My sisters Terrain gets no where near the rated highway mileage. And from looking at the Fuelly listings on both the Terrain and Equinox, no one else does either.

  • avatar
    George B

    One way I’d cheat if I was a car manufacturer would be to make the engine and transmission control computer start in a best fuel economy state and let the software adapt slowly to an aggressive driver. That way the test results for a new car would be from the hypermiling end of the spectrum while real world cars would become more responsive after they “break in”. I would also exploit unit-to-unit variation to make sure that the vehicle under test had the lowest possible friction losses and minimum weight for the test while still being a plausible representative vehicle.

  • avatar

    From what I have read Hybrid cars tend not to make their numbers but disels tend to do better in theirs, but to be fair I hear this from the disel group, any TDI people here or Hybrid people.

    • 0 avatar

      Hybrids face more variables for mileage. Ambient temperature effects how well the batteries perform, driving style has to be adjusted to get the most out of a hybrid moreso than a diesel, and the amount of city vs highway driving makes a huge difference.

      Even in city situations where most hybrids perform better the actual flow of traffic can change the results. More braking = more regenerative charging, but sitting in gridlock with the AC running drains the batteries without ever building up enough speed to be able to gain a meaningful regenerative braking charge.

    • 0 avatar

      We have a used 2005 Prius with 80K+ miles on it (I’m the 2nd owner). Right now the display says my tank average is 58.x MPG over the last 300+ miles.

      That’s on 100% gas, speeds up to 65 MPH, and with temps in the 90s so AC running on every trip.

      EPA sticker was 51 HWY, modified to 45HWY to compare vs 2008 sticker. So it beats the snot out of EPA in the Summer even with heavy AC use. In the winter it still equals the EPA numbers when using the heat but the MPG definitely drops in colder weather.

  • avatar
    Robert Fahey

    From the EPA website:

    “Manufacturers test their own vehicles—usually pre-production prototypes—and report the results to EPA. EPA reviews the results and confirms about 10-15 percent of them through their own tests at the National Vehicles and Fuel Emissions Laboratory.”

    There now, nothing to worry about.

    • 0 avatar

      Well that simply highlights the problem of self reporting. But I will note that most of the time manufacturers are honest. EPA may have it’s issues, but it is still massively better than taking personal anecdotes into this, or fanboys reporting on Fuelly…

      But Hyundai does seem to be fudging a bit here.

      Consumer Reports servers as a second comprehensive test of fuel economy. Consumer reports does all it’s own testing actually driving real cars they buy off the lots and using inline fuel meters (more accurate that fillup method or the “optimistic” on board computers). For these reasons I prefer CR to EPA, but as two sources of data, they can be used to sanity check each other.

      I have been comparing results between CR and EPA for some time and one solid pattern has emerged. Every single car I have ever cross checked, scores higher MPG on the CR Highway test, than the EPA Highway test. CR’s test is a straight 65 MPH highway run, while EPA includes stops/starts.

      There is one exception. The Hyundai Elantra. The only vehicle I have seen to score lower on CR, than EPA. It gets 40MPG on EPA and 39 on CR.

      Now these aren’t big differences, but compare results:
      Regular Focus Sedan: 43 MPG (EPA 38 MPG)
      Hyundai Elantra: 39 MPG (EPA 40 MPG)

      Ford has the typical difference, it gets 5 MPG more than EPA. Ford did not fudge.

      Hyundai fudged. They wanted to one up everyone saying all their Elantras get 40MPG EPA (while Ford/GM need special models).

      In reality though rated lower, the Focus sedan delivers better. Judging by the typical ratings difference between CR and EPA I would say the Elantra should be rated about 36MPG EPA, not 40MPG, a number chosen by marketing, not through actual testing.

      Again we are talking about small numbers. But Hyundai is using them in a marketing compaign to claim superior fuel economy to its competition, when in reality the reverse is true.

      EPA should prioritize testing an Elantra since it should be obvious Hyundai is playing games.

      • 0 avatar

        One thing about CR – they don’t publish (at least on the website for “non-plus” subscribers) the weather conditions at the time of the test, and exactly which gasoline (and the verified ethanol content) for each vehicle — when you’re in the lofty 40MPG range, these seemingly small differences could give the advantage to one or the other.

        That said, the Focus’ drivetrain outperformed the Elantra in acceleration and efficiency, but the DSG’s “balky” nature really hurt the Focus. Also, the Elantra actually bested the Focus by 3 MPH in the “Avoidance Maneuver”, which is a high-speed lane-change (so it’s not necessarily indicative of a car’s “sportiness”).

      • 0 avatar

        Considering the attention to detail otherwise, I would only expect they use the same fuel blend, and a similar range of conditions for their testing.

        I have never seen any inconsistencies in their testing to suggest otherwise.

        Hyundai fudged in this case so they could trump others on marketing numbers. Ford did their own testing to confirm their suspicions, but CR backs them on this one. If EPA actually tests the Elantra, there will be a significant correction.

        This kind of manufacturer gaming shouldn’t be allowed to stand as denigrates the value of EPA testing, which is partially dependent on good faith behavior from manufacturers.

  • avatar

    I would caution against ever trusting the dashboard trip mpg display. Rarely are they accurate. I believe part of this can be blamed on speedometers that are often a percentage point or more too optimistic. Our Honda Fit Sport AT is a great example of this. Its speedometer is optimistic by a few mph. I suspect this leads to the optimistic mpg’s shown by the trip computer. On the flip side, my Camry 4cyl LE AT’s speedometer is pessimistic (i.e. shows slower than actual if GPS is used) and returns better than posted mileage. I’ve had tanks where the Camry, with the larger engine (and far quieter and more comfortable ride) has returned calculated mileage of over 39mpg. This was while traveling at around 70 mph, west to east, so I probably had a tailwind. LOL. But I have been truly surprised by the Camry’s calculated mileage. 

    My Panther, 2005 Grand Marquis, was surprisingly fuel efficient, for being such a large car with a V-8 engine. The high gearing and low revs could deliver as high as 29 mpg at 70 mph. 

    My 2008 4Runner, while a very capable vehicle, was a what clues me in to what BS the dash trip computer was shoveling.  Before purchase, like an idiot, I believed what a lot of folks on forums were saying about their mileage with their 4Runners. They said they were getting low to mid 20’s. I then test drove one and it returned around 27 on the freeway at between 65-70. I thought I found the perfect vehicle. Of course, this was under perfect ambient conditions and terrain. Throw winter temperatures at it, a few hills, a bit faster speed. Then ditch the bs dash display, add a passenger or four and some cargo, and I regularly saw 14-17 mpg. In other words, I saw what the EPA promised or slightly less.

    So always do your own research and calculations. 

    As far as the subject of who might be gaming the system, I suspect Hyundai more than anyone.  I think they are good cars, but something doesn’t smell right.  The magazine comparison tests bear this out.

  • avatar

    I can’t say for certain but I’ve heard that the FIAT 500 consistently gets over its EPA numbers (30 city, 38 highway with a 33 average combined cycle). That is, the city mileage tends to be more like 32/33mpg, the highway closer to 40-42mpg with some hypermilers getting up to 45, perhaps more on the highway, but the average is 40-42 with the combined average being closer to 36-38mpg) if you keep your foot out of it at least some of the time. This car’s stubby front end tends to dramatically affect the mileage numbers if you shove the car along at greater than 70mph on the highway though.

    That said, when someone says I get XX mileage average, are they meaning combined average or is this their average highway mileage or what? I wish people would be more specific when reporting what they get and it’s not just here, but everywhere when the discussion of gas mileage comes up. I tend to assume they mean combined average but can never tell for sure though.

    I currently drive a flat nosed 2nd gen Ford Ranger with the 4.0L pushrod V6 and it at best gets maybe 27 on the highway, on a good day, going downhill with a tailwind so I think I average closer to 25 on the highway as long as I don’t go over 70. And this is a truck with almost 235K miles on it too. I don’t have positive numbers on this old beast and may be a bit optimistic, yes, it’s a 5spd manual, 2WD extended cab truck with a canopy so that plays a role in the mileage but it DOES seem to be decent for a compact truck none the less.

  • avatar

    I’d rather buy a car with EPA figures pulled from thin air than one geared and gamed for a badly modeled dyno test cycle.

  • avatar
    doctor olds

    EPA Mileage ratings are simply derived from the Federal Test Procedure emissions chassis dyno test. The actual amount of fuel used is not measured, but calculated from the tailpipe emissions numbers. The FTP offers a controlled and reasonable way to compare one vehicle to another on that specific drive procedure, not necessarily in the real world. If memory serves, the vehicle never exceeds about 50 MPH on the FTP. The actual numbers generated by the FTP were substantially higher than consumers achieved in real use. TTAC has covered in a general way the fudge factors used to generate the current, adjusted lower window sticker numbers. Real world usage is not constrained to the acceleration rates and maximum speeds of the FTP. Powerful, heavier cars can be quite a bit worse than the EPA estimates, as owners typically use the power to accelerate faster than the FTP.

    • 0 avatar

      Besides the above and one of the links provided earlier, it would be good for anyone reading the comments or this post to read and to know what actually goes on in the EPA tests.

  • avatar

    This could be coming from Mazda. They’ve outperformed many other makes with higher EPA numbers in CR, C&D, and MT fuel economy testing.

    I lost faith in EPA numbers when the Equinox’s were shown to be inflated. I no longer consider them a factor when it comes to evaluating vehicles.

    • 0 avatar

      The big Equinox buzz was the highway MPG. Which they actually achieve at CR:

      CR only tested the the AWD version rated at 29 MPG highway.

      CR got 30 MPG highway. It seems very likely that the FWD version would have matched the claimed 32 MPG. But still the number is close that there may be a slight bit of fudging. Usually a cars CR highway MPG will beat the EPA number by 10%.

      Still to this day they only car I have seen not make their highway numbers on a CR test is the Hyundai Elantra, which is Fudged at 40 MPG for marketing purposes. This is the one Ford complained about as well and rightly so IMO.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • Art Vandelay: I think the DMC-12 is pretty well served already. I saw a first gen Sable earlier this week and...
  • Art Vandelay: The Panther “heritage parts program” exists. It is known as your local pick n pull.
  • Art Vandelay: 240 has a pretty healthy aftermarket, though I don’t know if that extends to things like interior...
  • brn: At a minimum, TTAC (I like TTAC) should issue a retraction.
  • Art Vandelay: One kept me toasty on a cold night as it burned to the ground along side of the A1 Autostrada in Naples...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Jo Borras
  • Mark Baruth
  • Ronnie Schreiber