This question suggests using a tool or ruler to assess chain stretch to determine when to replace a chain. I have one such tool (Park Tool CC-3), and it tells me that my chain is not stretched even to the 0.75 level (is that %?).

My chain has at least 2,000 miles (~3,200 km) on it, so I am wondering if it should replace it just as a matter of practice (i.e., preventive care). I try to take care of the chain: wipe it down after rides, lubricate regularly (good advice here), etc.

So is it inconceivable that the chain is still in good shape after this many miles?

Is there any other reason to replace the chain other than elongation, assuming it otherwise appears to be okay?

3 Answers 3


The 2000 miles is an estimate of when the chain will stretch. Your chain may last longer than that.

As far as replacing it goes, I think it depends on your financial situation. If you can easily afford a new chain, I recommend buying one. A replaced chain means that you don't have to worry about your current chain stretching beyond the limit and damaging your bike.

If your budget makes buying a new chain difficult, you'll be all right to keep riding. Just check the chain stretch every so often to determine when it's getting near absolutely necessary to buy a chain.

  • Since Shimano recommends 800-1000 miles as the average life of a chain, where are you getting the 2000miles number? I'm asking because I see it recurring on this site in multiple answers from multiple sources, and I don't see it cited anywhere else. Serious question, not just nitpicking.
    – zenbike
    Jul 21, 2011 at 0:25
  • @zenbike You may be dissatisfied with this answer, but I believe people use 2000 miles because it's conventional knowledge, like getting an oil change for your car at 3000 or 5000 miles. These figures may not represent exactly when you actually need to replace these things, but they protect you against complications that may arise from ignoring their importance.
    – raabidfun
    Jul 21, 2011 at 6:53
  • 1
    I'm not dissatisfied, at all. I'm wanting to understand what seems to be common knowledge, but is different from what I've been taught. I was hoping there was a reference or a resource you could point me too. And I'm not saying it's wrong,. I'm oping to learn something new, is all.
    – zenbike
    Jul 21, 2011 at 8:34

If the tool doesn't show it to be stretched, and it's not rusted and there are no stiff links or other obvious problems, then there's no need to replace it.

It is a good idea to always keep a spare chain or two on hand, though, especially if you have an "odd" chain (eg, a 5-speed chain in these days where everyone else is running 10-speed chains). It's amazing how often a shop will try to sell you the wrong chain (or tube or tire) because they don't have the right one on hand.

  • Nice examination of your LBS. I've never, in 30 years of cycling, had a shop deliberately give me the wrong part. It makes no sense to do so, since it would create an angry customer, a return on your books, and guarantee twice the work for a simple transaction. If you have these issues regularly, perhaps we need to look at the common denominator in the transaction with these many bike shops that deliberately give you the wrong part.
    – zenbike
    Jul 21, 2011 at 0:09
  • 2
    It's not the totally wrong part, really, just not the right part. They'll sell a part that "will do" rather than go to the trouble of ordering the right part. Jul 21, 2011 at 1:33
  • Using a compatible but not perfect part, or waiting to order the exact part is something that's usually left to the customer to decide. But I don't know of any shop that I've been to that doesn't stock 5/6/7 speed chains. Bottom line: Does the part work? And did you inform the customer of the options?
    – zenbike
    Jul 21, 2011 at 3:04

First, chain wear tools are notoriously inaccurate because they rely on very specific conditions, including proper placement in the chain, and a user who is educated on how to use them. They can be used accurately, but commonly are not.Here is an excellent discussion of chain wear and how to measure it accurately.

Second, Shimano recommends 800-1000 miles on a chain as the point to start looking at replacing it.

Third, yes, it is possible that the chain is still in good shape after 3200km if you are a lightweight rider, who tends toward spinning rather than heavy gear use, and who maintains the bike well.

Last, if you're worried about it, replace it. There is no downside to replacing it early, and you've gotten your use out of it. Assuming it will not hurt you financially, of course.

  • I wouldn't say "notoriously inaccurate". Something like the Park CC-3 is pretty foolproof to use. And note that the error due to pin wear that's mentioned in that article will always make the chain appear more worn than it is, so there's no "risk" in waiting until the tool shows the chain worn -- worst case you'll end up replacing it too early. Jul 21, 2011 at 1:30
  • @Daniel R Hicks I said there was no risk in replacing it early. I did not say there was risk in waiting. And I said notoriously, because no shop guy I know trusts them. The difficulty is that the measurement is not isolated from the person performing the job. Meaning I can take that tool and get a measurement that shows worn, and my co-worker can take the same tool, on the same chain, and get a different measurement. I'm not saying he needs to replace the chain. I'm saying that the tool he's using to measure is known to have issues with accuracy. It can be used accurately, but commonly is not.
    – zenbike
    Jul 21, 2011 at 3:00
  • @Daniel R Hicks Also, if you reread my last two points, your comment is restating them. Was there something you disagreed with? Or are you just "not paying attention to who's posting"?
    – zenbike
    Jul 21, 2011 at 3:01
  • @Daniel R Hicks: The CC-3 often in my experience measures a chain with 3-5000 miles on it as unworn/new. That's not foolproof.
    – zenbike
    Jul 21, 2011 at 3:08
  • I've seen good bike mechanics use the CC-2. I user the CC-3, and it's extremely reproducible, so long as you don't force it. And there's no measurement tool made that won't give different measurements depending on who is using it (and how). Might as well throw out all rulers, eh? Jul 21, 2011 at 3:11

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.