I find myself avoiding steep/fast downhills. Not because I'm afraid I'll lose control, but because something might happen to my bike that will send me face-first into the pavement.

Are there any statistics on this?


By "crash due to bicycle failure", I mean a crash that would not have happened if something hadn't happened to the bicycle. This would include, for example, pounding your bike over bumps, resulting in the stem breaking. This would not include losing control because you pounded your bike over bumps.

By "crash", I mean an accident serious enough to be documented. If you fall over at a stoplight because you couldn't unclip, that obviously won't be documented and therefore doesn't count as a crash.

  • 1
    I think it's a good questin (+1) but here are rather large grey areas -- A puncture causing loss of control? Very worn brakes losing the last bit of friction material?
    – Chris H
    Dec 23, 2015 at 10:50
  • There are many crashes. I think a better question is injuries but I still doubt anyone is collecting statistics. Even if I have to go to the hospital and tell them it was a bicycle crash they don't ask if it was a mechanical failure. Personally I have had hundreds of crashes and only one a mechanical failure. I took you for more of a dare devil but it is good you are careful about speed.
    – paparazzo
    Dec 23, 2015 at 11:22
  • You might imagine that manufacturers keep data on under-warranty failures, but that is some way removed from what you're asking, and I can't imagine them ever making that information public
    – PeteH
    Dec 23, 2015 at 11:42
  • About 15 years ago I reviewed some of the best US statistics on bike accidents. I don't recall equipment failure even being mentioned. I'm sure the rate is quite low. Dec 23, 2015 at 14:22
  • Though I do recall one case of a roughly 15-year-old boy being injured (not seriously) when he attempted to hop a railroad track and the front wheel fell off. (This was before "lawyer lips".) Note that this failure would have been easily avoided by properly securing the skewer. Dec 23, 2015 at 14:24

3 Answers 3


There are some statistics available from Great Britain for Contributory factors for reported road accidents. Below are the 2014 figures for Pedal Cycles.

Contributory factor attributed to vehicle Number Per cent
Vehicle defects [subtotal] 267 2
Tyres illegal, defective or under inflated 10 0
Defective lights or indicators 54 0
Defective brakes 189 1
Defective steering or suspension 14 0
Defective or missing mirrors 0 0
Overloaded or poorly loaded vehicle or trailer 10 0

Note that only reported accidents where a police officer attended the scene are included in the statistics. This Wikipedia article has some details of the methodology used to collect the statistics.

The complete data set is available here. The data above is taken from the table RAS50005 - Vehicles in reported accidents by contributory factor and vehicle type, Great Britain, 2014.

  • Thank you, that is a really interesting table! It seems the brakes are the main component to worry about. Some brake failures may be also be behind the "too fast for conditions" and "junction overshoot" items. But it's also interesting to see that these numbers are very low, even taking into account substantial underreporting.
    – uUnwY
    Dec 24, 2015 at 9:56
  • 1
    Given that some of those items aren't required or generally fitted on bikes, the list must include motor vehicle defects causing bike accidents.
    – Chris H
    Dec 24, 2015 at 20:45
  • 1
    @ChrisH I was wondering about this too, and the full document isn't very clear about that. The full table has columns for other vehicle types where these rows are needed, so I guess e.g. the "mirrors" row applies only to other vehicles (and is 0 for bikes). On the other hand, there is also an item "vehicle door opened negligently" with 5 in the bicycle column. Surely that must refer to car doors?
    – uUnwY
    Dec 25, 2015 at 10:42
  • @ChrisH Actually looking at the full table, you might be right. There are some items that at least seem a bit strange for bicycles: "exceeding speed limit - 41" or "vehicle blind spot - 24" (although these could be coding errors for "too fast for conditions" and "failed to look"). There is one "pedestrian failed to look" and one "emergency vehicle on call" in the bike column (how many emergency bikes are there?), or "cyclist entering road from pavement" with 42 cases in the car column. These are a bit difficult to interpret, at least for me.
    – uUnwY
    Dec 25, 2015 at 11:00

The Royal Society for the Prevention of Accidents (RoSPA) collects statistics in the UK. I haven't read the documents in detail (you have to register to download...), but there is a summary page here: http://www.rospa.com/road-safety/advice/pedal-cyclists/facts-figures/

Mechanical failure doesn't seem to be common, it isn't even mentioned in the summary. Most injuries happen in a collision with a vehicle, so this is not the kind of crash you're asking about. The page says that, for adults, "16% of fatal or serious cyclist accidents reported to the police do not involve a collision with another vehicle, but are caused by the rider losing control of their bicycle."

Of course, "losing control" can mean a lot, including mechanical failures, and some of the collisions will also be caused by things like failing brakes.

As far as I know, there is a lack of data as bicycle crashes are normally not investigated in detail, unless there is a court case (but these are then usually collisions where guilt or liability has to be established).

From experience and general discussions it seems to me that the main mechanical failure that people have is in the brakes - the brake cable snapping, or worn brake pads (or the wrong kind of brake pads). The problem is that these issues may come without obvious warning - the brake may work well, but then suddenly fail completely when you need it. But you can reduce the risk a lot by good maintenance - check the brake pads and cables regularly for wear, and make sure both brakes are in good order so you always have one even if the other fails.

I've also heard stories of the front wheel coming off. However this is really rare and happened when they did a jump (e.g. jumping on a pavement, but they weren't going so fast). Not really on a road, as the wheel is pushed into the dropouts by gravity. The reason was really that the wheel nut wasn't tightened; so check that the wheel nuts or quick releases are fastened; again good maintenance reduces the risk. If you have a quick release and often leave your bike unattended, you may want to put a cable tie around the lever and the fork so that you can see if some joker fiddled with it.

If you have disk brakes, it is possible that the wheel is pushed out of the dropouts when braking, although this is more a design fault in some early designs and I think not so much of a problem any more.

Other types of mechanical failures (frame breaking, handle bar coming off etc) can happen in principle, but they don't usually happen so suddenly without warning, you'll usually notice the problem long before it's dangerous as the handling of the bike changes. For example, a crack in the frame may make the bike quite wobbly, but it doesn't suddenly disintegrate, and people have gone on for ages without crashing (or even noticing). I have once cycled for about hundred kilometers with several cracks in my rear rim.

So, mechanical failures can of course happen, but there are only a few types of failures that will be really dangerous (i.e. happen suddenly without warning), and much can be avoided by basic maintenance and generally checking your bike regularly.

  • I recall once seeing a bike with a complete fracture of the down tube at the downtube shifter bosses. (This apparently due to severe rust.) The bike was still in use (in relatively tame circumstances). Dec 23, 2015 at 14:26
  • 1
    Thanks for finding some stats, that's what was asked for :) Unfortunately, as you say, not common enough to be counted. Also, "quick release not done up, front wheel falls off" is disturbingly common, but it's user error rather than mechanical failure IMO. It's also a nasty crash because generally it happens soon after starting and people are taken completely by surprise, so facial injury is common. A friend of mine lost teeth, smashed up her face and lost a lot of skin off her hands doing that.
    – Móż
    Dec 23, 2015 at 19:54
  • I had a front wheel work loose. Clearly I hadn't done it up tight enough (hard to judge torque with pinhead skewers) . It made a disturbing rattle so I stopped and all was well. Had I gone over serious bumps I might have had a problem. Plus lawyer lips. So there are multiple layers of protection against what's normally user error anyway.
    – Chris H
    Dec 23, 2015 at 20:25
  • they don't usually happen so suddenly without warning usually not, but when they do, they are in full effect, especially true for aluminium and not so much for steel. Had a cheap alu stem break on me once. Knowing it was cheap I checked it for cracks about weekly. Nothing to see, yet one day I lift the front wheel just to end up with the bars in my hand and the wheel on the ground. Not good :P
    – stijn
    Dec 24, 2015 at 19:35
  • Disc brake forced ejection is only solved by Lawyer lips and Thru axles (on MTB).
    – mattnz
    Dec 26, 2015 at 4:16

No, there aren't statistics on this; "crashes" are not generally reported. Even if we we had all equipment-crash* data we don't have all human-crash** data.

*Crashes caused be equipment
**Crashes caused by cyclist error

EDIT to add for Frisby's sake:

The statistics that some guy compiled based on observations he makes by looking out his front window absolutely defeats my statement that the statistics don't exist. But that's obviously absurd. The point is that the data is not collected, it's not compiled, and no conclusions can be drawn - ergo no statistics which can be used to calculate risk.

  • 1
    @Frisbee I don't know where you are getting ur incorrect information but whatevs. The answer to the question is that that we don't know the percentage and it's because the stats either don't exist or aren't available. And yes, I can prove that we don't have complete crash data.
    – jqning
    Dec 23, 2015 at 19:09
  • Around 30 years ago some very dedicated cyclist set themselves up in several emergency rooms on the US East Coast for several months and interrogated every cycling accident that came in. I don't recall many of the details, but very detailed info on the accident cause was collected, and would certainly have included info on defect-caused accidents. So some information on the topic, albeit sparse, exists somewhere. (Note that medical privacy laws would no doubt prevent a repeat of that experiment.) Dec 23, 2015 at 21:51
  • @DanielRHicks That would be fun data to take a look at, the problem is that bike crashes don't always end in the ER. I'd go so far as to assume that most (by an order of magnitude) of crashes don't end in the ER. I'd hold very lightly to that assumption. That said, it is quite possible that OP is only concerned with the ER crashes!
    – jqning
    Dec 23, 2015 at 21:59
  • @jqning - You start splitting hairs pretty quickly. Is a skinned knee by a 4-year-old learning to ride a "cycling accident"? How about when I scrape my knuckles while placing my bike in a rack? Dec 23, 2015 at 22:43
  • @DanielRHicks which is why I asked in a comment to define crash.
    – jqning
    Dec 23, 2015 at 22:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.