Desert Highway

The Cost of Speeding: Save a Little Time, Spend a Lot of Money

One of our engineers is a guy named Ted, who spends a lot of time in the wilderness. He drives quite a bit, mostly on weekends to places like Yosemite National Park or Big Sur. He cares about the environment, especially since he spends so much time there, so he wants to be careful with his gas consumption. He also wants to save money. There’s just one thing: Ted drives fast.

At Automatic, we know staying under 70 MPH on the freeway can save gas and money, so we had Ted do the math to find out just how much. What he found out surprised even us. It turns out slowing down on the freeway just a little bit would save Ted over $550 in gas every year! Let’s dig in and see how.

Every car has an EPA-rated fuel efficiency. Ted’s 2011 Subaru Outback is rated at 29 MPG on the highway, but as you can see in the graph below the actual efficiency depends a lot on how fast he drives.

mpg-vs-speed-subaruAggregated data from over 2000 hours of driving across dozens of 2009-2013 Subaru Outbacks.


In the Outback, fuel efficiency peaks at around 50 MPH then drops quickly. Some cars are built to be more efficient than others – at lower or at higher speeds or both – but all cars follow this same basic pattern.

Ted’s average speed on the highway is 70 MPH, among the fastest Outback drivers using Automatic. He’s clearly burning more fuel at that speed than he could be, so he wanted to know whether slowing down would be worth the money saved on gas.

To answer this, he compared the gas cost and travel times of the fastest and slowest Outback drivers to the drivers who average 65 MPH on the highway. Ted’s in the fastest group.

gas-cost-vs-timeData from the 1000 highway miles Ted drove last month and normalized data from other Outback drivers. The fastest group spends $161 on gas and 14.4 hours driving; the median group spends $114 on gas and 15.4 hours driving; the slowest group spends $101 on gas and 18.1 hours driving. (Assumes $3.60/gal)


Driving an average of 5 MPH faster than the 65 MPH group, Ted saves only 4 minutes for every hour on the road but spends an extra $46 on gas every month.

A note about comparing time spent and money saved
At first, it may seem strange to compare monthly fuel savings to changes in travel time per hour of driving, but this is exactly the choice drivers have to make. Unlike the money saved on gas, which adds up over time, Ted can’t add up the few minutes he saves on every trip and use them later.

What about other kinds of cars? Whether you drive a hybrid like the Toyota Prius, a conventional Honda Civic, or a even a BMW 3 Series, which is engineered for better fuel economy at higher speeds, all cars follow the general pattern of decreasing efficiency above 70 MPH.

mpg-vs-speed-allAggregated data from over 8000 hours of driving for 2009-2012 model years of each car.

As you might expect, the Prius hybrid is the most efficient overall. The curve’s strange shape is because of its Hybrid Synergy Drive, which balances load between the gasoline engine and electric motor at different speeds. The Civic’s bump in efficiency at 70MPH is likely due to the combination of an overdrive gear and its VTEC system, which can improve fuel efficiency at high RPMs.

Despite these differences among cars, one point remains true: driving over 70 MPH will cost a lot of gas, but doesn’t save you all that much time. I asked Ted, and he’s not ready to join the slowest drivers on the road, but will start setting his cruise control to 65 MPH on the highway from now on. How about you?

Revised Feb 6, 2014: Corrected EPA rated fuel efficiency for Ted’s car from 24 MPG to 29 MPG and indicated the price of gas used in the analysis ($3.60/gal).


  1. No Spam   •  

    If Ted’s time is worth $45/hour and Ted saves 4 minutes every day (30 minute commute every way) on going the 5MPH faster, and there are 20 work days per month, then he saves 20 * (4/60) * $45 = $60 per month by driving faster. A rational choice over saving $46/month on fuel.

    “Unlike the money saved on gas, which adds up over time, Ted can’t add up the few minutes he saves on every trip and use them later.” Sure he can. Stay a 4 minutes longer at work Monday-Thursday, leave 16 minutes earlier on Friday. No one will mind.

    • Climber Dood   •  

      I know Ted, and the only thing he’s saving up his 4 minute chunks of #gangsterlife for are an extra dry tool workout or more cockpit time in a Mad Cat Mk II.

  2. Evan   •  

    It would be great to see data like this made available via an API. Both individual data from a user’s own vehicle logs as well as bulk data from all vehicles of a particular make and model could really make for some interesting insights.

    • Kris Erickson   •  

      I would like to second Evan’s request.

      I have been collecting data using an app called Vehicalc and I have downloaded the data and utilized it in some graphs I made in the past. However it would be very cool to but this product and let it log all that data automatically for me. How liberated is this data? Can I get a .csv file? Do you have an API?

  3. MPGomatic   •  

    It’s not just speed that determines efficiency. Driving style must be considered, as well. Setting the cruise control for a steady speed will be less efficient than using the road variables to your advantage.

    Adding a diesel vehicle to this comparison would be enlightening, even better if it has a highly advanced transmission. The BMW 328d xDrive is a prime example. I recently put it through the paces and the real world results were fabulous.

    • Ted   •  

      Hey MPGomatic–

      You’re totally right; there are a lot of variables that go into fuel economy– everything from the state of the car (tire pressure, roof racks, etc.) to environmental factors (temperature, how hilly it is, etc.) to driving behaviors (cruise control, “lead foot”, and so on).

      I’m actually very interested in cruise control and hope to write the next blog post on it. The information I’ve seen indicates it likely offers a significant improvement in fuel economy for most drivers under most circumstances, but I’ve definitely read it’s less efficient in some circumstances– most notably in hilly environments with a skilled and attentive driver.

      Unfortunately we don’t get sufficiently accurate MPG in diesels to do this kind of analysis, but I’m definitely hoping to in the future!

    • Joseph Z   •  

      True but cruise control does a better job at regulating fuel flow than your foot on the pedal does.

      If you set your cruise control in the city at 45 MPH average and when you go from a red light to a green light slowly up then resuming not increasing above 3,000 RPMs, you’ll notice the car kicks in the RPMs up to the point where it reaches 45 MPH then lowers the RPM ratio to the bare minimum to maintain the speed which is around 2,000 RPMs and only picks up if you go up a hill which raises it to 3,000 RPMs.

      If you do the foot on the pedal test, you notice that you go from 1,000 RPMs to 5,000 RPMs until you reach your steady speed of 45 MPH and you start cycling your foot between 2,000 RPMs and 4,000 RPMs.

      Also, we as humans tend to like to go above the speed limit versus a computer if the cruise control is set. Versus 45 MPH we tend to go all the way up to 69 MPH which is the borderline of the Automatic sensor to alert you you’re going over the safety parameter.

  4. Billy   •  

    An interesting piece of data the car companies already have but rarely share. Cool how the BMW achieves max efficiency at about 40 MPH and 68MPH. Ideal for local and highway driving.

  5. Tom   •  

    For the BMW, getting maximum efficiency at 70mph (112kph) makes perfect sense since a typical speed limit (when one exists) on highways is 120kph. Great engineering.

    For steady state driving, nothing will ever get much better than max efficiency over 70mph since drag force acts as the square to speed. At 80mph, you car acts as if it’s driving through a thicker medium than air, at 100mph that thickness practically doubles!

    But rather than focus on cruising speed (unless you normally drive 100, in which case SLOW DOWN you might as well dump the cash out of your wallet!) the biggest difference is the rate of acceleration. All day long I see people rush up to red lights, then race back to 40 mph, repeat, repeat, repeat. This is the equivalent mpg of going 100mph at steady state! Rolling up to reds so they turn green before you fully stop and smoothly accelerating back up to speed will make a huge difference in MPG vs slowing down 10mph on the highway.

    • Caleb   •  

      IIRC, Those window stickers are based on specific tests, and vehicles in the US are engineered to pass those tests as best as possible. In Europe, they’re geared towards actual driving – specifically, highway driving. I remember reading an article some 20 years ago that claimed BMW/Mercedes get significantly better highway fuel economy on their European sold vehicles than the US ones due to gear and final drive ratio differences. Don’t know if still true.

      It would be nice, in today’s world, if you could order your car custom geared for predominantly city or highway driving. Or as the 328i shows above, perhaps some manufacturers are already making 2 sweet spots.

      I also believe, at some point, drag becomes a significant factor. My BMW (528i) may get that double-peak in the 70 MPH area, but my SUV may just drop like a rock, not due to engine or gearing, but due to drag alone. There’s a good flat stretch I’ve been meaning to drive both cars back and forth on at different speeds using cruise to generate this graph for myself. Maybe it’s time to do it.

  6. Ryan   •  

    Those efficiency vs. speed curves are very interesting. As the user base and amount of data collected grows, will be be able to see similar cars for more makes and models of car? I’d love to see how my car’s efficiency changes relative to speed. Perhaps it has some variety of “sweet spot” like the Honda Civic that can help me maximize my fuel economy, since the bulk of my driving is on the highway.

    • Ted   •  

      Hey Ryan,

      Glad you liked them! We’re definitely planning on doing this for more cars. Even cooler, one feature I’d like to eventually add is a “personalized” fuel-efficiency curve that shows how your car deviates from the car’s “overall” curve for all Automatic users– this could provide insight into vehicle-specific deficiencies, such as low tire pressure, etc.

      • Ken   •  

        Great idea, would be a huge help.

        This could be taken further to +/- the 70 MPH rule on a vehicle-by-vehicle basis.

  7. Luke   •  

    I love this data! Do you have it available for more cars? I’d love to know what the curve looks like for my 4-cylinder Accord. In future updates, will I be able to produce a graph like this for my own car. (I’m an Automatic user -love it.)

    Also, do the curves take into acceleration into account at each speed? I know when I am going 20 – 30 mph, I am often changing speed, whereas at 65, I’m usually “cruising.”

    • Ted   •  

      Hi Luke!

      We’re working on ways to deliver this kind of information on a per-user basis via the app. I think it’d be fascinating to see how individual users’ curves differ– it might tell you all kinds of interesting stuff like low tire pressure, etc.

      These curves are aggregate data for all driving at that speed, so they do reflect the impact of driving behaviors (i.e. the MPG shown between 20 and 30 is probably significantly lower than the car’s constant-velocity MPG, whereas the MPG around 65 is closer to its constant-velocity MPG, because as you noted people tend to cruise at highway speeds).

  8. Nick   •  

    Can you publish or tell us where we can get the curve for our cars?

    • Ted   •  

      Hi Nick!

      We haven’t published other curves yet, but we’re working on giving users insight into this aspect of their car in new versions of the app. Stay tuned!

  9. John Parriott   •  

    I have a question. Because I am mechanically inclined and love tech tools that keep me in the know of what my vehicles are doing I have a question as to your products capabilities. You have stated that this product can check codes and even clear them. However you do not mention about monitoring things like fuel rail pressure, intake pressure, fuel filter restriction, O2 sensor readings. Does your module allow someone to check on these on a constant basis, or only when the check engine light comes on? The reason I ask is because I can get a bluetooth scanner that does check on current readings for less than $15 at sears, or on amazon. It won’t tell you the location of your car or call the police in the event of an accident, but that is why I have onStar, and take mental/visual cues on where I parked. If you are trying to do for the car world what apple has done for the music industry than I highly suggest a little less focus on panoz and a little more focus on the tech and how it can help humans be smarter…..not lazier. It does look good and from the pictures I see for the app, very user friendly. Just seems like you cut yourself short and could afford to go a little further.

  10. Tom Hansen   •  

    I would like to see more technical details about exactly what data is captured by the device. If I could choose, for example, to log all data so I could crunch it myself later, that would be a big selling point for me. As the other poster indicated, there are Bluetooth OBD-II interfaces already available that will let you access all the data.

    SUGGESTION: create a plugin API that would allow developers to access all the data for various purposes. The Automatic would then serve as the user-friendly platform for lots of things you guys haven’t thought of.

  11. Jeremiah   •  

    Our radio show would love to feature this device on our weekly Tech Talk radio show and you tube broadcast. Anyway we could work something out? Please check out our website and show and let us know.

  12. Brian   •  

    Maybe I missed where you addressed this, but it appears you are comparing and/or combining the results from Outbacks with different engines. You can’t do that. If Ted’s 2009 Outback is rated 24 MPG on the highway, that means he has the turbocharged 2.5XT model. The XT is much rarer than the non-turbo models. And those non-turbo models get better gas mileage. Your chart shows data for 2009-2012 Outbacks and you use it to illustrate the MPG by speed for Outbacks like Ted’s. But the multitude of engines and their ratings for the models in those four years makes that chart meaningless and deceiving.

    • Ted   •  

      Hi Brian– you’re quite right; that’s a typo. My car’s highway EPA is 29 MPG, not 24– we’ll get that fixed.

      You’re also right that engines make a huge difference! For this article, I lumped together standard (non-turbo) 2.5 and 3.6 liter engine Outbacks, because I found driving styles outweighed engine characteristics for fuel economy.

  13. Guest   •  

    Go as slow as you want, just don’t do it in the fast lane. 😉

  14. Terry O'Connor   •  

    There is a lot more to obtaining good gas mileage than the normal driving speed. In 2005 I purchased a new Hoda Civic Hybrid. Just “after” signing the papers the salesmen asked me what gas mileage I expected to get. I voiced that I would expect at least the advertised mileage of 48 city and 47 highway at which point they balked. They said I should expect around 40 mpg average depending on my driving habits. Since I commute approx. 60 miles per day and my main reason to purchase a hybrid was for good mileage I took that as a challenge to get at least the advertised mileage. During break-in I drove the car reasonably easy and was achieving around 44 mpg. I knew I could improve my driving habits and began by eliminating jack rabbit starts, backing off sooner prior to stop signs and signals and timing stop light and stop signs so that I limited coming to complete stops to a minimum (still making a complete stop at all stop signs and red signals). My mileage immediately began to climb and now for the last 10 years and 150K miles later I average at least 51 mpg per tank and have a best of 56 mpg for over 500 miles of driving on a single tank of gas. This is still driving 65+ mph on thehighway, going over a grade to and from my office and doing normal around town driving.

  15. Glenn   •  

    Doesn’t make sense. Travel 1,000 miles at 70 mph takes 14.3 hours. Travel 1,000 miles at 65 mph takes 15.4 hours, so 0.9 hours (54 minutes) are saved at the faster speed.

    Either I’m missing something or the study is using two different prices of gas.

    From the graph, at 70 mph, Outback gets 29 mpg, at 65 mph it gets 31 mpg. Therefore, for 1,000 miles driven gas consumption is 34.5 gallons at 70 & 32.25 gallons at 65. $161 cost of gas/34.5 gallons = $4.67/gallon cost of gas at 70, but $114 cost of gas/32.25 gallons = $3.53/gallon. So the calculations are way off because two widely different prices for gas are used.

    If the cost of gas in each case is $3.50/gallon, then at 70 mph the total cost is $120.70 and at 65 mph it’s $112.90, a difference of $7.80. Anybody’s time is worth $8/hr.

    • Ted   •  

      Hi Glenn!

      We used a gas price of $3.60, which is generally what I pay at the pump.

      Unfortunately, you can’t use the chart to extrapolate a precise MPG for a given driver at a given speed. It’s aggregate data over all drivers of the car, whereas the fuel economy numbers for users reflect their different behaviors.

      Driving faster doesn’t just put you in a different part of the curve, it also affects the rest of your driving pattern– your overall speed becomes a lot more variable as you lane change to pass people, brake due to traffic, etc.– and so your efficiency drops more than the curve would indicate.

      So the curve reflects aggregate data, but the “save a lot” is based on the true travel times and fuel consumption for the fastest/slowest/median drivers. Does that make sense?

  16. E Drews   •  

    An interesting article. I drive a good number of miles every week, and I’m a firm believer in the use of a cruise control. One doesn’t need all of the fancy apps to figure out some of the information. Use some simple record keeping:
    – Calculate your fuel economy every time you fill the tank. Track the miles per tank on a trip odometer and note the gallons purchased when you fill the tank. The calculation for MPG is simple arithmetic. Keep your records for a few months or a few years. Make adjustments to your driving to see if your fuel economy numbers improve based on that adjustment.
    – Use your cruise control. The cruise controls of today are far better than those on vehicles of 5 – 10 years ago. Hills really don’t make that much of a difference to the cruise control. Watching the speed on a GPS, or something relatively accurate, will tell you how much change there is in the speed of your vehicle. My observations show the set speed is maintained within 1 mph.
    – Does speeding save time? It is easy to check. At least some GPS units have elapsed time as part of their data collection. They also have average speed. My back-of-the-envelope, in-my-head calculations seem to indicate that if you are stopped in traffic on an interstate class highway for more than 30 seconds on your trip, it is unlikely you will be able to make up the time. Watch the ETA on a GPS unit when you start your trip, and watch it move forward when you’re stopped in traffic. It doesn’t take too much to make that ETA move later, but it takes a whole lot to bring it back to where it was when you started your trip.
    – The length of your trip is going to depend heavily on your average speed. Is your average speed improvement better served by increasing your maximum speed,, or increasing your lowest speed? Getting stopped by law enforcement is going to kill your average speed. My thought is to concentrate on keeping moving, rather than moving fast. Driving in a manner that keeps your foot off the brake (back off from the vehicle in front of you) helps a whole lot too.

    Happing motoring

  17. Amy Gurley   •  

    hmmm… I’ve gained enough real world, average data through Automatic to try reducing my speed in the morning from my avg of 68 mph down to 60 and see if i save much. I gain very little benefit to arriving early to work… I just sit in my car and read. On my way home, traffic is lame. I’m lucky if I go over 60 mph at any given length of time.

  18. Mike   •  

    The methodology used in this analysis appears to be seriously flawed. I just crunched the numbers using your chart above for fuel mileage, the national average annual miles driven, and the current average California gasoline prices (based on where Ted likes to spend his free time, I assume he lives in California).

    First, the assumptions:
    From the chart above, the curve shows 32.8 MPG @ 55 mph, 31.5 MPG @ 65 mph, and 29.5 MPG @ 70 mph
    1000 miles/month (from data provided)
    Current average California gasoline price is $3.565/gal

    Now for the results (on a monthly basis, with the above assumptions):

    Speed (mph) 55 65 70
    Driving Time (hrs) 18.2 15.4 14.3
    Fuel Used (gal) 30.5 31.7 33.9
    Fuel Cost ($) 108.7 113.2 120.8

    As one would expect, there is clearly savings associated with driving slower. However, driving 70 mph vs. 65 mph only costs an additional $7.60, while saving 1.1 hrs. Driving 55 mph vs. 65 mph saves only $4.50, but it costs the driver an additional 2.8 hrs in driving time. Comparing 55 mph to 70 mph, the faster driver pays an additional $12.10/month in fuel expenses, but he has 3.9 more hours each month to spend with his family, watch TV, or do whatever it is he enjoys doing. On an annual basis, this means that for only $146, Fast Ted spends 46.8 hours LESS sitting in his vehicle than Slow Ted. That’s almost two whole days! I dunno…maybe Ted doesn’t get along well with his wife and enjoys the extra time in his car. :)

    Based on my analysis of the data, the only way Ted could achieve $46/month is savings by driving 65 mph instead of 70 mph is if he is paying $21.37 per gallon of gasoline. Not very likely.

    • Ted   •  

      Hi Mike!

      I just replied to a similar comment from Glenn (above) but the gist is that you can’t extrapolate a precise figure from the curve. It might be reasonable if you drove a thousand miles on flat ground at a constant speed at the two different speeds, but in reality, driving faster results in a much less efficient drive cycle overall (i.e. more acceleration and deceleration, which wastes a lot of gas)

      So I derived the curves from aggregate data from all Outback data, then compared drivers based on their average speed– which inevitably includes a slew of inefficient behaviors associated with higher speeds, and reflects more than just their location on the aggregate curve.

      and I’m not married, but maybe I could use my time savings to improve my methodologies 😉

      hope that helps!


      • Mike   •  

        Thanks for the response! I suspect the majority of the savings comes from driving more conservatively, as opposed to driving at slightly slower cruising speeds. Those who speed also typically drive more aggressively – quicker acceleration, hard braking, etc – all things that severely impact fuel efficiency. Speaking from experience here. :)

    • Ted   •  

      Oh, also, we used $3.60 for gasoline– should’ve noted that in the article, sorry!

  19. Adam   •  

    Not only is the methodology flawed, if you look at the chart, the most common vehicles, the standard sedans, have a much later peak in the chart. The Outback is among the earliest. Unfortunately for your conclusion, it looks like if you take the Honda Civic or BMW, your savings are severely diminished, and you’re spending pennys to save dollars of time. Infact, on the Honda chart from your own data, it sure looks like 70MPH is significantly more efficient then the 65MPH. A more accurate statement to display larger would be:
    Driving an average of 5 MPH faster than the 65 MPH group, Ted driving a Honda saves 4 minutes for every hour on the road but more importantly, saves an extra $200 on gas every month. (although you used different gas prices, as others have noted)

    • Ted   •  

      Hi Adam!

      High-speed fuel economy is dependent on a lot of things, but the vehicle’s aerodynamic drag is a major factor. You’re right that sedans typically have lower drag (and thus better fuel economy at high speeds) than larger cars.

      If you’re interested, I’d suggest you take a look at the wikipedia page for aerodynamic drag in cars:

  20. Patrick Egan   •  

    Look don’t get me wrong cause I really like the Automatic but gas savings at higher speeds generalization is going to get you some bad press. The reason some states are now even raising the speed limits back to 70 is because of the findings that the drive 55 to save gas program started in the 70s was baseless in the first place and with the modern technology in most vehicles is unfounded. I buy that the with the Subaru test vehicle you may have valid results but making this statement for all vehicles is going to backfire. My Audi set on cruise at 78 gets me from LAX to SFO cheaper than doing it at 65 based on hard dollars and time spent on the road. And looking at the new Lexus IS with a 8 speed gear box I would be very skeptical.

    But… having said that I would love to see the empirical data from spit out by Automatic just to be sure.

    So long story short. Until you have all the details for all vehicles, PLEASE do not make blanket statements like that it just diminishes the reputation of a great product.

  21. Dan   •  

    Or you can get a 5/6 speed manual, get boosted MPG and decide how hard your engine is working. If you’re going 45mph at 2k rpm at 4th, if you put it in 5th you’ll save a lot of gas if you cruise at 1.5k rpm. Same with freeway speeds, 6th will put you around 1.5k rpm if you’re cruising at 65mph.

  22. Wc19805   •  

    In general the statements are true. Its the folks that love speeding that take any info to justify their behavior.
    I used to speed to work too in an 18 year old Cadillac. With a 4.9 liter engine, 2.73 final drive my motor turns about 1,200 rpm at 60. Fuel economy runs about 28-30 mpg. At 70 it drops to about 23~24 mpg. This is according to an actual readout on the vehicle. Once I lost my job & started making a whole 80% less than I was making I started running 62~65 on the highway. Basically I became “unimportant” overnight. So the extra few mins an hr make no difference. But I can do quite a bit with an extra $46 a month. One thing the article don’t mention is driving over mountains makes a big difference because of the barometric pressure & temperature changes. Nothing u can do about that one. So granny driving in Florida is going to get much better mileage than a granny driving in Denver, Co. But in general if u run 60 instead of 70 you will save gas.

  23. Dave   •  

    I think you all missed it.

    The cost of the gas is not the only cost for Ted. There is is also
    the extra time it take to pump the more gas he is using per month…. how many extra stops is that?….not clear.
    But then there is definitely more wear on the tires (150 x 4), as tires wear more at higher speeds, and probably an extra visit to the brake shop somewhere over the life of the car. If Ted is very unlucky there will be a visit to the Judge and the speeding ticket, all of which will set him back many months of time and many dollars.

    Way better to drive slower.

  24. John   •  


  25. Annabela   •  

    I love this!

  26. Test   •  


  27. Ted   •  

    Bode, here’s a holistic suggestion: try using Automatic 😉

    Actually though, if you’re interested, the EPA has a pretty good page with some basic tips to get you started. I’ll be exploring some of them in a data-driven way, and down the road we’ll be looking into giving rich, personal insights to our users based on their driving patterns.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>