“If my satnav shows I’m going 4/5 miles an hour slower than my car speedometer says, which do I believe?” Asked by Tabitha, September 2012

Having recently bought a used car, Tabitha began to notice that there was a consistent discrepancy between the speed showing on her speedometer and that on her portable satellite navigation unit.  At most speeds, the speedo reading was several mph faster than her satnav unit was indicating.

Many other drivers have reported the same thing over the years, and this situation is actually common to nearly every car on the market.  But why is this the case?  Surely car manufacturers can make their speedos accurate to the precise mph or km/h you are travelling at?

How a car speedometer works

The Car Expert explains why your car speedometer may not be accurateSpeed is the measurement of distance over time.  But a car speedometer doesn’t actually measure how fast you travel from Point A to Point B.  Car speedos usually work by measuring rotation of the car’s driveshaft, axle or wheel.  They then use some basic maths to extrapolate that rotation and determine how fast you are travelling.  It’s a very similar concept to a bicycle speedometer.

However, if the diameter of the wheel/tyre alters, the extrapolation calculation will be incorrect.  For example, the diameter will increase if you put new tyres on the car (more tread, which wears down over thousands of miles) or increase the tyre pressure.  This means that, for each revolution of the wheel, the car is travelling further, meaning your speed is greater.

If the diameter decreases (eg – worn tyres, less air in the tyres, a different brand of tyre with slightly different dimensions, more load in the car weighing it down and compressing the tyres), then the car will be travelling a shorter distance for each revolution of the wheel, therefore you will be going slower.

Margin of error in a car speedometer

The differences in wheel diameter resulting from the above circumstances could be tiny (maybe a few millimetres), but at 30mph your car wheels are rotating 6-7 times every second, so it can quickly make a difference of a few miles per hour.  This margin for error is taken into account in how the law is applied, and how manufacturers calibrate their car speedos.

How a satnav speedometer works

Is the speed on your satnav more accurate than your car speedometer?Satellite navigation units (either portable or integrated into the car) calculate your car’s speed by measuring actual distance travelled over time using GPS satellite tracking.  They repeatedly locate your exact position on earth via satellite and calculate how far you have travelled, then divide by the time it took for you to travel that distance.  Satnav accuracy is determined by satellite signal quality and is unaffected by your car’s tyres.  Many satnavs are unable to account for changes in vertical direction, so may be less accurate if you are travelling up or down a steep hill.  They are also inherently more accurate at higher speeds, as a larger distance over time reduces rounding errors, but a satnav will usually be much closer to a car’s true speed than the speedometer.

Some factory satnav systems will also use data from the car to integrate with the GPS signal to improve overall accuracy.

The law for car speedometers in the UK

The UK law is based on the EU standard, with some minor changes.  A speedo must never show less than the actual speed, and must never show more than 110% of actual speed + 6.25mph.  So if your true speed is 40mph, your speedo could legally be reading up to 50.25mph but never less than 40mph.  Or to put it another way, if your speedo is reading 50mph, you won’t be doing more than 50mph but it’s possible you might actually only be travelling at 40mph.

To ensure that they comply with the law and make sure that their speedometers are never showing less than true speed under any foreseeable circumstances, car manufacturers will normally deliberately calibrate their speedos to read ‘high’ by a certain amount.  As your satnav is not the designated device by which a car’s speed is measured, it does not need to incorporate any fudge factoring.

Was this article helpful? Tweet about it so others can get get help too!

If you have a question for us to cover in a future article, head over to the #AskTCE page and it will be answered soon.  If you found this article helpful, please take the time to share it with your friends on Facebook, Twitter, LinkedIn, Google+, StumbleUpon or Reddit.  Thank you :)
How does your satnav speed compare to your car speedometer?

Join the Forum discussion on this post

Stuart Masson is the owner and editor of The Car Expert. Originally from Australia, Stuart has had a passion for cars and the car industry for over thirty years. He spent a decade in the automotive retail industry, and now works for automotive editorial agency Immediate Network. Stuart writes most of the articles on The Car Expert, offering impartial and independent advice about buying and financing a car.


  1. Would a digital speedo be more accurate than an analogue dial-type speedo? Or do they have the same fudge factor?

    • Hi Connor, it shouldn’t make any difference, as the calculation to display the speed is the same for both analogue and digital speedometers. These days, they often use a digital signal from the wheel sensors anyway and convert it to an analogue dial display. So yes, they both have the same ‘fudge factor’.

  2. How do we know that the satnav is accurate? I thought that the satellite signals were deliberately designed to be less accurate than possible for “security reasons”?

    • As far as I know, you are correct in the level of accuracy for civilian GPS units (like satnavs) being lower than military applications. But if you are moving at 30mph, you are covering more than 13 metres per second, which makes it possible for a satnav to measure your speed accurately. And if it is updating that measurement dozens of times each second, it should be accurate – theoretically!

      • The system known as ‘selective availability’ which introduced a timing error for civilian GPS use was turned off a number of years ago – mainly because it had become quite easy to circumvent the introduced error using differential GPS (a fixed installation in a known location on the earth can be used to calculate the timing error and pass it on to any other GPS capable of receiving that extra info).
        As far as I’m aware, there’s no longer any difference in the accuracy of civilian vs military GPS.

  3. I’m surprised that manufacturers haven’t come up with a better way of measuring actual speed than measuring rotation. Surely it couldn’t be that difficult to bounce a tiny radar or sonar off the ground to measure real speed?

  4. Air pressure in the tyre makes no difference to the measured speed of the vehicle. The length of the outer perimeter of the squashed circle is the same as the circumference of the circle as would be on a properly inflated tyre. One rotation of the whole wheel requires the entirety of the outside surface to travel along the road and since that hasn’t changed, neither has the number of revolutions of the wheel, gearbox or measurement system.

    Fitting different size tyres will affect it though as you have stated because then you’re actually changing the length of the rubber on the outside of the tyre.

    Additionally, some sat-nav speedometers don’t account for the fact you’re travelling up or downhill, so their speed indicates only the horizontal displacement of the vehicle, and Pythagoras tells us that the hypotenuse (road) of a triangle must be a greater distance than the horizontal (where vertical is height), so the speed on your sat-nav may be slightly lower than your actual speed in this case. For example: If you drive up a 30 degrees hill and your sat-nav says 40mph you may be going closer to 50mph.

  5. UK-driven trucks have calibrated speedos which is fine but causes problems in certain circumstances. Roadworks usually have 50mph so us HGV drivers drive close to the speed limit if we can, but hey that car in front using the second lane is also travelling at 50mph, so us HGV drivers in lane one catchup to car we also start using lane two, the car will not move faster or move into lane one, so us HGV drivers might give car driver a quick flash which might get the car driver to move into lane one. Then you might get lucky and get backup to 50mph. But car driver thinks you are speeding and then does an illegal move which is apply their brakes for NO reason. ALL manufacturers should stick to the truth and issue calibrated speedos.

  6. The ECU of my car reports a speed that is within 0.01 mph 0f the speed shown by my sat-nav. But the speedo shows 70mph when the satnav reports a speed of 65mph.

    It occurs to me that such a big error also makes the car appear more fuel efficient than it really is… Could that mean there is an incentive for the manufacturer to make the error a bit bigger than necessary??

    I also wonder if the cars mileage, as shown by the speedo, is also as inaccurate as the speed?

    • Hi Joseph. If a modern car is using the wheel sensors to measure rotation speed, the gear ratios shouldn’t matter – the wheels are still turning at the same speed even if the engine revs are higher or lower. To illustrate, if you are driving your car at 30mph in 2nd gear, 3rd gear or 4th gear, your ratios are very different but the speedo is still accurate.

  7. How come when a car publication (I can't remember which one) tested several second hand cars, they found the Alfa 156 was reading 30 mph and the car was actually doing 33 mph. It was a completely standard car. As the car went faster the speedo started to read faster than actual speed.

    • Hi Lyndon. Always difficult to assess the validity of used car results, as you don't know the history of the vehicle in question. Unless you tested a large number of the same model and found a consistent problem, it's impossible to blame an inherent problem with the car. I used to own an Alfa Romeo 156 and never had this issue.

  8. So I take this screws it up for owners of classic cars who have worn out 40-50 year old speedometers and cant get replacement ones. Mine is 1998 and (Calibrated) Digital so I'm ok just another money making scheme.. now people are buying more Hybrid / Electric cars and not spending as much on fuel they need to get the loss back some how..

  9. If the circumference of the tyre is smaller then the wheel will have to rotate more times to cover the same distance so speed will appear higher. Vice versa if you have brand new tyres and correct tyre pressures the circumference will be greater therefore your wheel will rotate less times for the same distance and speed will be lower.
    You stated it the other way round.

    • Hi Steve. Your maths is absolutely correct, but you are calculating backwards. Distance travelled is an unknown for the car, not a known.

      Wheel circumfererence (known, but with fudge factor as described above) x wheel rotation speed (known, measured from axle or wheel sensors) = vehicle speed. So a larger circumference for a given rotation speed will yield a higher vehicle speed.

  10. The indicated speed on my 2007 (8th Gen) Civic reads 4.5mph over what I'm actually travelling at tested against trucks from our own fleet !
    I understand perfectly what Les was saying above re: travelling through roadworks. The muppets double up and are actually travelling around 43mph in a 50 zone and point blank refuse to return to Lane 1. Have a guess what the hypocrites then do as soon as they get out of the AVERAGE speed camera zone ?
    The mind boggles.

  11. How does this inaccuracy affect the recorded mileage on the vehicle? If the speedo is over reading by 10% does this resulting a10% increase in miles and hence more services and money spent at a garage?

  12. Hi everybody. This article has been getting huge numbers of referrals from Facebook over the last few weeks, which is absolutely fantastic, but Facebook doesn't allow website owners to track where exactly on Facebook the link has come from. If the link is on a public page, can somebody please point me in the direction so I can find it?

    Many thanks, stuart.

  13. I am aware that EU law requires that speedometers never under-read but they can over-read by up to 10 per cent + 6.25mph.

    I have a part-time delivery job and the van's speedo regularly and consistently over-reads by 20 per cent as calibrated against my satnav

    Is this legal? What implications does it have on the vehicle's MoT check assuming the examiner was to find it (and as yet the last three tests that I have been aware of haven't). Would he check for that and if so, how?)

    • Hi Trevor. It's not part of the MOT test, as far as I know. If it's over-reading it's not really a problem as you're not likely to be going over the speed limit. You could probably send the van to the dealership to have it fixed, but unless it's still under warranty it will cost you.