I'd have thought this would have been asked already but I didn't see it. Is driving the speed limit safer on a highway? By "highway" I mean high-traffic areas, such as 4-lane divided thoroughfares.

Having commuted for some time, I've noticed that people who clearly drive at excessive speeds seem to be dangerous. However, more often than not the danger I see is from people who are driving slower than the prevailing rate (where the prevailing rate is generally more than the speed limit). This seems more dangerous because it disrupts the flow of traffic, irritates drivers and people behind the slow driver end up tailgating.

Is there any evidence to suggest that:

  1. speed limits on highways improve the safety of people driving on those highways?
  2. people driving the speed limit and those around them are safer than those travelling at the prevailing rate of traffic - even where that prevailing rate is faster than the speed limit?

EDIT It is noteworthy that there has been news recently that Google's self-driving car is programmed to go faster than the speed limit, apparently because it can be safer in some scenarios.

EDIT In line with this is a comment from the video Why you shouldn’t drive slowly in the left lane by Vox, which states “Research has shown that the strongest predictor of an accident is variance from the average”. Which seems obvious (and hence this question), but I would like to know what the paper is and whether it's been debunked.

  • 1
    Aren't the main contributing factors in serious accidents 1) excess speed, 2) alcohol, with 3) not wearing seat belts, 4) being young and invincible, and 5) distraction by cell phone use, coming up behind? Yeah, people driving the limit may irritate younger drivers, but in a sense they are doing them a favor. Commented Jun 2, 2011 at 21:23
  • 7
    @Mike: More speed is more energy, and more energy is more problems if an accident happens. However, if everybody's going the same speed, accidents are less likely, and so there may be fewer deaths with everybody going about 70 mph as opposed to some at 70 and some at 55. I don't know how to test that. Commented Jun 3, 2011 at 2:48
  • 6
    You are asking two different questions here, and although they're related I'm not sure they're close enough to make one good question. Commented Jun 3, 2011 at 2:50
  • 2
    @Mike I was under the impression speed differentials were one of the major contributing factors. If everybody is driving 70 it will probably be fine, but it you have somebody sitting in the middle lane at 30 that is unexpected so a group of cars coming up behing that are doing the speed limit need to take avoiding action, if they fail and accident is caused
    – Ardesco
    Commented Jun 3, 2011 at 13:49
  • 2
    Despite what any poll or survey might say, the number one cause of accidents are people not paying attention, or simply being a lazy driver. Yeah, I'm looking at you Mr. I change lanes without using my turn signal because it's so hard to reach two inches down to activate it. Afterall, I should be reading your mind and know that you want to change lanes, right?
    – crush
    Commented Mar 14, 2014 at 16:11

4 Answers 4


Higher average speed will lead to more accidents. This is because braking distance doesn’t just double from 100 km/h to 200 km/h. In fact, it quadruples. It’s nonlinear, but people think it’s linear. Insufficient following distance is the second-biggest cause of accidents (the biggest is excessive speed). Translate this link.

You might have thought that big traffic jams are mainly caused by accidents. But in fact, it’s the difference in speed that will on higher speed increase non-linearly probability of abrupt braking, insufficient car-to-car distances and longer braking distance of cars. Modern traffic systems try to decrease the human factor as much as possible in order to reduce braking distance. Then you can increase speeds. This is because several cars will behave like a railway train with fixed distances between train carriages.

The second point you mention to adapt to avg. speed (> speed limit) to drive safer is kind of misleading and mirroring this linear thinking (esp. for very old drivers with reduced reflexes), as people will tend to choose non-linearly too small distances to the car in front with increasing speed. Trucks in the right-hand (slow) lane are very unlikely to cause traffic jams. This is because they behave mostly like the "road train" that was described in the hyperlink in the previous paragraph.

We do a lot of traffic research here in Germany. This is because some of our speed-limit signs are computer-controllable, depending on traffic density, average speed, and traffic-jam messages from drivers. These limits get adapted temporally on highly frequented highways, but only downward from the maximum speed limit. Also you probably know of new car features like "Autonomous cruise control system", that will use computer control to control the speed and distance to cars in front to you.

So of course if the speed limit is 100 km/h, one shouldn’t drive 50 km/h in the rightmost lane. Small speed-limited two-wheeled motor vehicles (max. 50 km/h) are forbidden on German highways. As far as I know, in Europe the highway speed limit of around 120 km/h and an advisory speed limit around 80 km/h is the best compromise of fast traffic and human factor. In Germany we have, in some areas, no general speed limit, since we are car fanatics and have a strong lobby :) However, most European countries have a general speed limit of around 120 km/h.

  • 10
    Being in police work for many years, I have been interested in the various aspects of traffic safety. Read a fascinating book last year called "Traffic, or why we drive the way we do." Dealt as much with the psychology of driving as actual safety measures. A lot of cherished notions turn out not to be well-tested or supported. IMO, speed considered alone is rarely a factor. When you throw in inattention, tailgating, failing to adjust for conditions... It's perfectly safe to drive at 200 mph bumper-to-bumper, NASCAR guys do it all the time. Till something goes wrong.
    – M. Werner
    Commented Jun 10, 2011 at 19:55
  • 6
    @Tormod No, the autobahn doesn't generally have different speed limits per lane.
    – Ruben
    Commented Jun 10, 2011 at 21:00
  • 3
    This answer could do with some more references to support some of the claims.
    – Oddthinking
    Commented Apr 16, 2012 at 11:28
  • 2
    Certainly in the UK the statistics actually do not support the claim that higher average speeds lead to more accidents. There is a strong correlation with the damage done in the event of an accident, but the major causes are lack of concentration, lack of training, lack of experience and tiredness. see safespeed.org.uk
    – Rory Alsop
    Commented Aug 31, 2012 at 9:50
  • 3
    @Rory Alsop lack of concentration, lack of training, lack of experience and tiredness all of these are worse at higher speeds.
    – Stefan
    Commented Aug 31, 2012 at 12:49

This is a near-miss answer: it addresses only 60 km/h roads, not highways. Nonetheless, it addresses some of the attitudes to driving described in the question, so it may shed some light on the answer:

For 60 km/h speed-limits:

  • Driving faster than the speed limit is associated with a huge increase in the chance of being involved in an accident involving a casualty.

  • Driving slightly slower than the speed limit doesn't make much difference.

  • Driving very slowly (e.g. 40 km/h in a 60 km/h zone) is associated with a moderate increase in the chance of being involved in an accident involved a casualty.

This is illustrated in Table 4.3 of a 1997 report from the NHMRC Road Accident Research Unit of The University of Adelaide: Kloeden CN, McLean AJ, Moore VM, Ponte G, Travelling Speed and the Risk of Crash Involvement, Volume 1 - Findings.

Their headline finding:

In a 60 km/h speed limit area, the risk of involvement in a casualty crash doubles with each 5 km/h increase in travelling speed above 60 km/h.

Graph of relative risk

Correlation is not causality. Correlation is not causality. Correlation is not causality.

  • 2
    But the stats leave out key issues of some relevance. Are they based on the absence of other road users? Would a similar pattern exist if the road were designed for 120km/hr travel? I imagine that going very fast on a road with a 60km/hr limit where most users obey the limit is very unsafe but doing the same on an uncrowded motorway might not be.
    – matt_black
    Commented Apr 16, 2012 at 23:16
  • 4
    Yes, the roads would be safer if there were no other road users, the roads were designed for faster travel, people were trained to drive faster, all roads were straight, and people were made of titanium. :-) I can't see the point of your hypotheticals. I will grant (and do in the first sentence) that stats for 60 km/h roads are not necessarily 100% applicable for 100 km/h roads - I don't think they can be totally dismissed though.
    – Oddthinking
    Commented Apr 17, 2012 at 0:59
  • 5
    One thing to point out is that these stats point to risk of casualty (ie death) and not just risk of injury or even just risk of a crash. Of course there are going to be more deaths 70 or 80 km/h then at 60, because the car is travelling at a much higher speed, and is carry a lot more energy (E = 1/2 mv^2). Also cars are pretty good at protecting us at low speeds but they aren't designed to protect you as well in highspeed crashes. Crash tests in the US are only done at 35 mph (56 km/h) because the chances of surviving a high speed crash are very low.
    – Kibbee
    Commented Sep 1, 2012 at 0:37
  • What about selection-bias?
    – user1873
    Commented Jun 19, 2014 at 22:33
  • 1
    @user1873: I am trying to understand what you mean. If you mean only poor drivers elect to speed (or drive significantly lower than the speed limit), you have a fair point. This is one of the reasons I have reminded people I haven't shown causation.
    – Oddthinking
    Commented Jun 20, 2014 at 1:54

Several of the other answers here talk about speed, energy and braking distance. But that's not what the question asked.

The question asked whether it is safer to drive at the speed limit, and makes reference to other cars driving both faster, and slower than this figure.

The answer to this is that there's no objective answer. Speed limits are a legal construct, and have no intrinsic relation to physics or engineering.

By far the biggest factors in road safety are:

  • Paying due care and attention to those around you, and being alert for potential hazards, while taking mitigating action to avoid them, and;
  • Driving at an appropriate speed for the conditions at the time.

The appropriate speed depends on many factors, for example the density of the traffic; whether that section of road is straight and clear (or narrow and winding); presence of junctions; presence of parked cars; weather (snow, ice, fog, rain, or a clear dry day); driver experience; driver reaction times; and braking ability of the car.

This is not an exhaustive list. However it illustrates the complex factors which will come into any conceivable equation when calculating "What is the safest speed". Clearly, this is a question which can not be answered objectively.

With regards legal speed limits, you tend to find the same speed limit applied to the same road - regardless of whether one section is straight and clear, but another section has many junctions, or is a common place for cars to park at the side. Furthermore, most speed limits don't change on a daily basis according to the weather. This is why on a clear, dry day many legal limits feel absurdly low - whereas in snow and ice, on the same stretch of road the legal limit may be dangerously fast.

Part 2 of the questions asks whether it's safer to drive at the legal speed limit even when the prevailing speed of traffic is faster. Again, the answer is: "it depends". If the prevailing speed of the traffic is excessive, and the traffic is so dense that stopping distances are compromised (i.e it's actually an unsafe speed), then it would indeed be safer to reduce your speed to increase the distance between your car and the one in front (even at the expense of the annoyance of drivers behind you.

On the other hand, if the density of traffic is relatively light and there is plenty of stopping distance, then it's safer to go with the flow. The reason for this is because if you're driving significantly slower than the rest of the vehicles around you, you're causing them to swerve around you; change lane; and otherwise alter their behaviour by making them take additional manœuvres. At high speed, this can create more inherent danger than simply factoring in the energy calculation.

To recap: overall, higher speeds involve more energy, and therefore a reduction in the overall energy on the roads means that in the event of an accident, there is less chance of damage and loss of life. However this is unrelated to the "legal speed limit"; and there are many other important factors to road safety. Drivers who spend their time with their eyes fixed on the speedometer are generally paying less attention to the road around them.

  • 1
    Despite being a legal construct, speed limits do have a relationship to physics and engineering. Thankfully, policy is sometimes informed by science.
    – De Novo
    Commented Nov 2, 2018 at 16:51
  • 1
    Do Novo - care to explain your logic? Care to explain why my narrow residential road, with loads of parked cars, has a legal limit of 30mph; while the nearby main road (wide and open) also has exactly the same legal limit of 30mph? There's no physics-related reason for that. It's all just about a nice round number being decided upon by the government and slapped onto all roads in the area, regardless of their features. It's a political construct: not a physical one. Commented Nov 2, 2018 at 21:40
  • 1
    No need to invoke logic. It just happens to be done that way. Legislative bodies consult traffic engineers, who provide recommendations based on physics and engineering principles. Policy, meet science :)
    – De Novo
    Commented Nov 2, 2018 at 21:52
  • 1
    "No need to invoke logic" - then no point in continuing this discussion. Commented Nov 2, 2018 at 21:54
  • 3
    "Speed limits are a legal construct, and have no intrinsic relation to physics or engineering." This is an empty argument. Which side of the road you drive on is purely a legal construct, and yet driving on the wrong side of a highway is empirically much more dangerous.
    – Oddthinking
    Commented Nov 3, 2018 at 0:13
  1. That depends on the specific highway. For example, in British Columbia, it seems as if increasing the speed limit actually causes accidents to go down (source - there's a video near the end of the page that explains why this is so). You have probably heard about the German Autobahn, where in some stretches there's no posted limit, and yet it's famously safe to drive on (in 2011, there were 5.6 fatalities in German roads per billion vehicle-km - source).
  2. No, people are safer when everyone is driving at the same speed, going in the same direction. The first link above has several sources to back up this claim.
  • 3
    Welcome to Skeptics. 1: Changing the speed limit is a separate question. 2. Could you quote the part that makes this claim? Even better, chase it up and quote the original sources.
    – Oddthinking
    Commented Mar 21, 2014 at 2:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .