Several general questions have been asked regarding how long a bike or bike frame will "last", and the gist of answers seems to be that a bike can last quite a while of the replaceable components are replaced and the bike is reasonably well maintained.

But what about the inherent structural integrity of the frame itself? Is there such a thing as a risk of sudden frame failure or disintegration after some level of use? Do bikes ever fall apart, or develop a risk of breaking or fracturing along weak points? One of my bikes has more miles on it than the car I'm turning in — and I wonder about the integrity of the car!


Yes, frames do fail even if they're not crashed or ridden excessively harshly. The only way to mitigate this problem is to inspect the frame regularly during your maintenance and cleaning. Look for cracks. When riding, pay attention to creaks and squeeks and always find the root cause (it could be a crack).

Keep in mind that:

  • A bad weld could hold fine for years before coming apart.

  • Unchecked rust can eat through steel tubes, either from the outside-in or the inside-out. Once it starts, you have to keep on top of it or it will eventually take the frame. Rust also seems to have "good taste" in tubing by preferring the pretty Italian stuff to, say, gas-pipe schwinn.

  • Metal fatigue can form cracks that grow slowly until the frame suddenly snaps (I've seen this personally where the chainstay meets the dropout, and on crank arms).

Aluminum is the worst culprit for fatigue cracks. Aluminum seatposts and steerer tubes also have a tendency to develop aluminum oxide blooms that may permanently jam the seatube/steerer-tube in the frame.

Carbon typically won't survive a bad-enough crash. I've heard of long-term delamination issues, but never saw one personally. AFAIK every carbon frame "dies" in a crash or mishap that damages the frame to the point where the owner no longer trusts it. Anyone have stories about what happens to carbon that is not crashed?

For a concrete example, below is my own failure from a few years ago. It happened as I was cresting a very steep hill. I heard a loud "pop" and didn't realize what happened until I looked around and saw the chainstay separate from the dropout with each pedal stroke. This is a steel bike and was totally repairable, BTW. I didn't crash as a result (but neither was I going fast). Note the little bit of rust visible at the top of the crack. It indicates that the crack had been there for some time before the total failure. enter image description here

  • 1
    In the case of cracks or rust, there usually is a sign something is going wrong. Cracks take time to grow-- months or years, before catastrophic failure. The problem is that you might not know a crack is there until it causes a failure. They're hard to see. That's why it is a good idea to inspect the frame and root-cause each creaking sound. AFAIK, there is no MTBF (mean time between failure) for a bike frame. It would depend on too many factors (loading, random material imperfections, etc) and it would be statistical and thus not really useful for making individual decisions. – Angelo Oct 4 '11 at 20:03
  • 1
    That's pretty much what I was looking for (not that I thought to ask it that way): whether there were MTBF stats. But it sounds like, overall, it's not even enough of an issue that it's even discussed much. Certainly, it seems there's nothing like a recommendation to "discard after X miles", which one would expect if the issue were real (assuming the industry is properly regulated)—it's just hard to believe that something can be that durable (I change my phone every six months)! – orome Oct 4 '11 at 20:14
  • 1
    I haven't yet seen a CF delamination on a bike, but have seen it on cars and airplanes, so it's definitely something that can happen. – Brian Knoblauch Oct 5 '11 at 17:59
  • 1
    Also your point about carbon I feel is little too general, most of the DH frame manufacturers are now moving to carbon for the strength it gives. They're made in such a way that the frames don't weigh much less, but are much stronger than an equivalent weight of Al due to the fact that the properties are tunable by the way the laminates are assembled. – cmannett85 Oct 6 '11 at 10:47
  • 1
    @cbamber85, you're absolutely right. Carbon is fantastic and in usage dominates all competition bike frames for good reasons. It does what it is designed to do very well, but crashes put forces on parts of the frame in a way that no design can accommodate. – Angelo Oct 6 '11 at 16:37

What you are referring to is sometimes known as catastrophic failure, when a frame fails under load. While this does happen from time to time, it's relatively rare. There are tales of defective carbon frames failing spontaneously while riding, but these seem to always be friend-of-a-friend stories. Most frames that do "fail" do so because of traffic accidents or other damage. And, as Gary points out, the bottom bracket (and mounting eyelets) are likely areas to wear.

Frame material also plays a part in this. Conventional wisdom in cycling circles is that steel is the most durable - it can be bent back into shape and even re-welded. Aluminum frames can't. And carbon fiber frames have a reputation for being delicate; if one of those breaks, you throw it out and get a new one, since any fixes will be unreliable.

While there is some truth to this, steel can be damaged to the point where it would be unsafe to fix and re-use. And carbon fiber is more durable than folks assume. But evaluating a damaged frame is often more trouble than simply replacing it. There are pro wrenchers who can do this kind of evaluation based on experience, but without putting the frame through an x-ray machine can you really, definitively say that a frame is safe to put your weight on after it's been bent and re-welded?

In general, yes, the frame is the most durable part of the bike. But it, too can be damaged beyond repair.

  • 2
    Just a note on the failure of carbon frames... My friend / team-mate broke his carbon framed mtn bike clean in half during a race. We were wondering why his lap was taking so long, when here he comes jogging out of the woods with his bike over his shoulders. Broken in 1/2, the 2 sides just held together by the cables. The manufacturer gave him a replacement frame in the end, but it does happen! – rally25rs Oct 4 '11 at 19:13
  • 2
    Another note on carbon frame failure: bustedcarbon.com – alxndr Jul 18 '13 at 16:23
  • Actually carbon is very repairable and unlike steel the repaired section is stronger after the repair than it was before the repair. It kills me when I see idiots in shops tossing out perfectly repairable carbon frames. Of course these are usually the same people who tell customers that aluminum is stiffer than steel when the opposite is actually true. – ChrisL Jul 13 '15 at 19:20
  • Tales of defective carbon frames are not alway "friend of a friend stories". There was a cases recently of a deadly crash when a carbon fork disintegrated (admittedly whether you see the fork as part of the frame is debatable, but it's a similar problem), see road.cc/content/news/… – Stephan Matthiesen Sep 19 '16 at 17:17

Just some notes on spotting cracks in metal frames:

Look for cracks in the the paint. These can be quite small even ~5mm long for instance.

Look at the stress points of the weld - that is usually just at the edge of the weld bead where the metal is most thin. Bottom brackets take the most stress of the bicycle but rarely fail because they are the most supported. Dropouts are most common areas of failure over time in my experience. Most crashes are from the front and will show up in the fork or where the head tube meets the down tube. Those are the spots to look.

Aluminium can snap suddenly. I've had this happen on handlebars. Usually steel will bend before it breaks and is much safer in that capacity. People who speak about carbon being durable are either fanboys or salespeople. Everyone knows carbon can break dramatically.

Front dropouts, fork crown and handlebars are really the ultimate safety check spots to watch. failures there can kill. Rear dropouts breaking will be bad but not super dangerous usually.

Fancy lightweight road bikes are the most susceptible. Frame failure on a steel commuter bike usually happens long after the bike has been junked for other reasons. If you are really worried about frame failure watch that front end, don't use aluminium or carbon and you will be fine riding even ancient bikes.

Sometimes slightly large or heavy individuals >200lbs will have more issues with frame fatigue than the rest of us as most bikes are designed for smaller riders.


Yes - frames are subject to wear, tear, aging and fatigue. The primary area of concern or failure on most frames is at the bottom bracket. On both metal and carbon frames this is a point of high stress and flex. But any weld or joint is subject to stress and may fail over time.

You should periodically inspect the whole frame, paying particular attention to and weld-points or joints. Look closely for signs of paint or weld cracking as well as any rust on steel frames. The best way to do this is to hand wash the frame and inspect it as you go. If you want it to last even longer apply a layer of bike or car wax your frame every six months or so. Wax makes it harder for dirt and mud to stick to your frame, and will make the next clean easier.

That said, if you make it a habit to clean and inspect your frame, are an 'average' rider (not excessively heavy or riding aggressively off jumps and drops), store the bike properly and care for it well it could conceivable outlive you.


I have seen a few old steel frames that failed. In one case the down tube cracked in two at the point where the down-tube-mounted shifters were attached. In another case the crack was at the junction between down tube and bottom bracket. I've also seen seat stays separated at the seat tube.

But all of these failures occurred on bikes that had quite obviously been left out in the rain for years (if not decades), and all of them were still being ridden, even with the broken frame. The traditional steel "diamond" frame is extraordinarily robust.


My Giant Explorer’s frame had a sudden catastrophic failure a couple of weeks ago. The duralumin frame split in two without warning across both tubes as I was dismounting, with the two wheels completely separating. It is a mere 9 years old, I bought it new, no-one else has ever ridden it, and it has not been mistreated in any way. Being a roadster type with a built-in carrier it should be designed to carry a load. It has had a fair amount of use, but only on roads and cycle tracks, and mainly for shopping, travel or pleasure. The tubes fractured at what is clearly a weak point, where the cables entered the tubes. This should be a warning to other owners of this model to check the frame in this area, especially the bottom tube. I was only relatively lightly injured, but it could have been far worse, even fatal, had I been in traffic or travelling at speed.

  • In my opinion having cables enter the frame is always a weak point and a sign of poor design (except in the case of very high end bikes where there is a lot of material added inside - even then its going to be a stress point). I've seen cheap steel frames buckle at the insertion point several times. – rusl Mar 9 '16 at 5:54
  • FYI Giant warranties their frames for life. You should have got a new equivalent bike out of it for no cost, or had that retail cost subtracted from the cost of a higher-spec bike. – Criggie Apr 21 '17 at 5:23

I managed to break a steel frame about 2 years after buying it. That did involve a crash a year after owning it and then riding it to destruction though. I noticed after the crash (into a ditch) that the geometry felt shorter and tighter. I just carried on riding it though. In hind sight that might not have been the most intelligent thing to do.

  • That's a concern. I've ridding this thing for long enough, and treated it badly enough, that I wonder when and if it will just give up the ghost. – orome Oct 4 '11 at 20:43
  • As I say, there were some clear indications that the frame was damaged. The geometry changed, i.e. I could tell that the wheels were closer together afterwards. If I had closely inspected the frame I suspect I would have seen flaked paint and an indication of the damage too. On the day it finally did give up the ghost it felt funny, groaning and generally not feeling right too. I didn't come to any harm when it snapped in this instance although it was unrideable afterwards. Basically you have to ignore a lot of signs to actually let that happen. – Colin Newell Oct 5 '11 at 7:55
  • @2u2, my alluminum frame took about 10 days of commuting, all the time groaning and flexing, to break. A large crack was growing about 2-3 mm a day, until it went all the way around the upper tube. – Vorac Nov 22 '13 at 10:54

I've seen three bicycles with what appear to be stress fractures at the bottom of the seat tube and top of the bottom bracket shell. A bike nut friend with more metallurgical knowledge than me tells me that this is caused by work hardening -- which is the same thing that causes a paper clip to become brittle after repeated flexing. Now that this has happened to my trusty chrome-moly Fuji Gran Touer SE frame after 20+ years of daily use, I have a suspicion that the habit of mounting and dismounting on/off the left pedal while rolling is something that probably puts lots of extra stress/flex on this joint.

  • 1
    Work hardening is much more likely with aluminum than steel. You hardly ever see it in (properly engineered) steel. – Daniel R Hicks Jul 12 '15 at 13:33
  • 1
    I'd also note that the bottom bracket is the place in a standard steel frame which is most prone to internal rust. – Daniel R Hicks Jul 12 '15 at 14:00

Years ago I had a steel frame (Trek 700) break at the seat stay and dropout. I think this may have been caused by flexing due to the fact that my rear axle had previously broken. The bike had been well used for at least ten years.

FWIW I also once broke a JIS crank spindle, maybe also due unusual forces from a loose crank fixing bolt.

protected by Gary.Ray Apr 21 '17 at 12:29

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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