Support road.cc

Like this site? Help us to make it better.

feature

Do you need a new chain? Find out the easy way to tell

A worn chain can wreak havoc on the rest of your transmission. Read our expert guide to learn how to tell that it's time for a new one

Your bike chain will gradually wear with use and will need changing from time to time in order to maintain your drivetrain’s performance. Worn chains shift poorly, wear sprockets quickly, and sometimes break. 

So when should you change your chain?

“For chain replacement we do not state 'every x kms' as this is not possible,” says SRAM. “Chain wear is based on multiple factors including maintenance (clean/lube), use conditions (water/mud/sand), user shifting patterns and overall drivetrain condition (cassette/ chainring wear).”

Cross-chaining: is it really all that bad?

Campagnolo agrees.

“It is difficult to pin down an exact number to kilometres due to the fact that riders come in different weights and sizes, ride differently, shift more or less frequently, develop more or less wattage, ride on flat or hilly terrain, clean or nasty conditions, take care or leave their chain dirty… all of which create large variables in just how much wear and tear is created,” says Campag's Joshua Riddle.

“It can vary between 3,000km to 8,000km generally speaking, but it could be less or even more in some cases.”

KMC X10.93 Chain

First, you need to replace your chain when you spot any damage (a deformation or crack). You should also check your chain regularly to see if it has worn to the point that you need to change it. 

There are several ways to check whether your chain has reached this stage?

Measuring with a ruler

You can check for chain wear with a ruler. It’s a little easier if you go with imperial measurement here because one complete chain link of a standard chain measures 1in.

Start at one link pin and measure 12 complete links. You need to put some tension on the chain to be accurate.

Chain Checking inches - 1.jpg

On a standard new chain, 12 complete links will measure 12in, but when a chain is worn the 12in mark of the ruler won’t quite reach the relevant link pin.

If the distance from the 12in mark to the centre of the link pin is less than 1/16in, your chain is fine, but if it gets to 1/8in (or 2/16) mark it has gone beyond the point at which it needs replacing.

If you use metric measurements, it’s easiest to measure 10 links.

Chain Checking cm - 1.jpg

On a standard new chain, 10 complete links will measure 25.4cm.

If the distance from the centre of one link pin to the centre of the link pin 10 complete chain links away is up to 25.5cm, your chain is fine, but if it gets to the 25.6cm mark it has gone beyond the point at which it needs replacing.

If you want to get more accurate, it's usually advised that you replace chain designed for 10 or fewer gears when it has lengthened by 0.75% – so when 10 links measures 25.59cm – and that an 11-speed or 12-speed chain is replaced when it has lengthened by 0.5% – so when 10 links measures 25.53cm. 

If you have a one-speed or two-speed bike, replace your chain as it reaches 1% wear - so when 10 links measures 25.65cm.

Figures like these are obviously very difficult to gauge with a normal ruler, which is why we'd advise the use of a chain wear indicator.

Chain wear indicator

A chain wear indicator, sometimes called a chain checker, is an inexpensive gauge that does exactly what its name suggests. 

Park Tool Chain Checker - 1.jpg

This is Park Tool’s CC-3.2 chain checker with an RRP of £9.99, although you’ll find it cheaper if you look around. Other brands offer similar instruments.

You hook the curved end into the chain and if the gauge tip on the other end fits completely into a chain link, the chain is worn to the point it needs replacing (one side measures 0.5% wear, the other side measures 0.75% wear). 

Measuring by eye

There’s one other simple method of checking for chain wear. 

Chain Checking by hand - 1.jpg

KMC advises, “If you do not have a gauge to test the chain’s elongation you could roughly check by putting the chain on the outer chainring and lifting up the chain from the middle of the chainring. If you can lift it more than half a link, the chain or chainring, or both, are probably worn.”

The chain in the picture is almost new so minimal lift from the chainring is possible.

Check out our advice on how to replace a chain. 

Mat has been in cycling media since 1996, on titles including BikeRadar, Total Bike, Total Mountain Bike, What Mountain Bike and Mountain Biking UK, and he has been editor of 220 Triathlon and Cycling Plus. Mat has been road.cc technical editor for over a decade, testing bikes, fettling the latest kit, and trying out the most up-to-the-minute clothing. He has won his category in Ironman UK 70.3 and finished on the podium in both marathons he has run. Mat is a Cambridge graduate who did a post-grad in magazine journalism, and he is a winner of the Cycling Media Award for Specialist Online Writer. Now over 50, he's riding road and gravel bikes most days for fun and fitness rather than training for competitions.

Add new comment

73 comments

Avatar
Mat Brett replied to nortonpdj | 7 years ago
0 likes

nortonpdj wrote:

One simple thing not given enough prominence in the article: the cleaner you keep your drivetrain, the longer the chain will last.

 

The scope and aim of this article is clearly set out in the headline: When should I replace my chain?

And in the standfirst: Here's how to know when it's time to fit a new one

There's plenty more information we could have included about chain longevity, chain maintenance, how to change your chain, and so on, but that's not what this article is about. The article is an answer to an often asked question.

For advice on cleaning and lubing your chain go to http://road.cc/content/feature/164006-how-clean-and-lube-your-bikes-chain

Avatar
Crashboy | 7 years ago
1 like

This pre-emptive maintenance is a bit scary to me and my wallet.  TBH I usually change my bike before I change the chain!

Admittedly, I've never done hyper miles, but I've never felt the need to check or replace a chain unless it was totally ruined (i.e snapped / siezed / super rattly / a really out of shape link).  I would only even think of checking it if there was a problem with my shifting.  

I've also never felt the need to  replace a cassette or chainrings unless the teeth on them have done similar or it won't shift properly / makes terrible noise and all other solutions have failed (ie. adjusting mechs / new cables etc)...

For context, my commuter bike does Roughly 100 - 150 miles per week all weathers and I clean & lube the chain every 3-4 weeks  if it's lucky when I sluice the bike down (fortnightly in the winter or if squeaking starts).  

Most exotic drivetrain I've had is Tiagra..is that a factor? Do the more expensive parts wear more easily because they are lighter / thinner / different quality and therefore more delicate perhaps?

What am I missing out on by not changing this more frequently? I'm concerned that this sounds like a big important maintenance deal when I presumed a chain was...the most boring part of the machine and pretty indestructible!

 

Avatar
CXR94Di2 | 7 years ago
0 likes

Chains I just change once a year, I do about 2000-3000 miles.  I use KMC because their link is easy to take off/put on.  I just fitted a new Shimano chain to my in-laws bike(he bought the chain), what a faff, those brittle little pins.  You need a new one everytime, the chain has to come off.  

 

Those who are concerned about wear, just use a steel rule to measure across 12 links in inches.

Avatar
Boombang replied to Paul__M | 7 years ago
0 likes
Paul__M wrote:

Rotating chains sounds a good idea, but when I got my spare out of the draw amazingly it had started to rust, so storage needs to be well lubricated if the garage is at all damp. That's a bit messy for me, so am going the 'disposable item' route.

I degrease chain on removal and relube, then put into a Ziplock bag. A quick spray of whatever combo of 2 letters and 2 numbers you got laying about, seal up and stick in drawer - I write rough mileage on the bag too.

All that's left is give the chain a good wipe down and refit.

Avatar
timtak | 7 years ago
0 likes

> This is Park Tool’s CC-3.2 chain checker (link is external) with an RRP of £9.99, although you’ll find it cheaper if you look around.
They are less than £1 from China. https://goo.gl/jxl7Z5 (no affiliation).

Avatar
Anthony.C | 7 years ago
1 like

I reckon a lot of people are replacing chains prematurely, those tools don't really measure chain 'stretch' they measure roller wear. The only reliable way to measure is with a good ruler. Having said that I just replaced a chain because it was dropping in easily on the 0.75 % side but I got a few thousand winter miles out of it.

Avatar
WolfieSmith | 7 years ago
3 likes

Always worth checking the jockey wheels too. Here's three years worth of wear with a chain change once a year on a summer bike. As you can see - the chain chews away pretty fast.Probably 6000 miles worth.

Avatar
. . replied to CXR94Di2 | 7 years ago
2 likes

CXR94Di2 wrote:

I just fitted a new Shimano chain to my in-laws bike(he bought the chain), what a faff, those brittle little pins.  You need a new one everytime, the chain has to come off. 

You can buy KMC links separately and use them to reconnect Shimano chains.

Avatar
mike the bike replied to . . | 7 years ago
0 likes

. . wrote:

CXR94Di2 wrote:

I just fitted a new Shimano chain to my in-laws bike(he bought the chain), what a faff, those brittle little pins.  You need a new one everytime, the chain has to come off. 

You can buy KMC links separately and use them to reconnect Shimano chains.

 

And clever little things they are too.  My three bikes are all equipped with a 'Missing Link", it makes removing the chain a 20-second doddle.

Avatar
Applecart | 6 years ago
0 likes

I really recommend Ultegra chains. I run 105 groupset and the generic (105?) chain my Cube came with snapped randomly. All the shop had was Ultegra and it felt immediately brilliant - silky smooth shifts and more solid power transfer. Not expensive, and I'm more likely to keep it clean for that good transfer + offset the cost.

Avatar
Applecart replied to Paul__M | 6 years ago
1 like

Paul__M wrote:

Darkerside wrote:

One of the down sides of a recumbent. Chain swaps need three of the damn things...

Feels a little less disposable at £60+ a pop.

 

That's another reason my High Racer is a 'Sunday' bike. Must get it off the trainer now the clocks have moved.

 

If the chain is 3 times longer, it will last 3 times longer. Think about it  - the average wear on each link is 3 times less as it is in contact with a sprocket under force 3 times less frequently over its lifetime. Therefore the price is the same.

Avatar
Sub4 | 6 years ago
2 likes

Campag. I change chains shortly after .5%generally get 5  + chains before cassette rejects the new chain. Those cassettes are pricey! Record chains do last very well.

Shimano chains seems to wear v quickly & cheap ones seem to be at 0.5% from the box.

KMC is my chain of choice in the workshop.

As everyone has said, clean & lube regularly & they last a long time. Ignore them & watch them melt in a few weeks.

ps, if I had a quid for every time a customer said "that chain's nearly new, it can't be worn!" when the 1% goes through, I'd have lots of quids.

Avatar
996ducati | 6 years ago
0 likes

Having snapped a chain on a winter sportive I always carry the tool and links just in case now.Personally, for some reason my chains seem to wear diagonally (as opposed to streching) which the measuring tools dont always pick up so easily (cross chaining?) and at the rate of about two per year on 3000/4000 miles.

Avatar
mathcore | 6 years ago
2 likes

7800 km on my izumi super V and I ve just passed the 0.5% of stretch .  , the zen chainring still black on the teeths , she doesnt show any sign of wear .  JAPAN !

Avatar
barbarus | 6 years ago
1 like

Replace preferably before it looks like this:

 

Avatar
nbrus replied to Anthony.C | 6 years ago
1 like

Anthony.C wrote:

I reckon a lot of people are replacing chains prematurely, those tools don't really measure chain 'stretch' they measure roller wear. The only reliable way to measure is with a good ruler. Having said that I just replaced a chain because it was dropping in easily on the 0.75 % side but I got a few thousand winter miles out of it.

Yes, chains don't actually 'stretch' ... it is roller wear that is being measured and those chain neasuring tools work very well and are much simpler and more precise that trying to keep 12 links of a chain straight, in tension, lining up a ruler and trying to take a reading for stretch. However, I would not pay £9.99 for a chain wear tool ... I got mine from On-One for £2 some time back.

If you want your chain to last, then use a good wax dry lube.

Avatar
Anthony.C replied to nbrus | 6 years ago
0 likes

nbrus wrote:

Anthony.C wrote:

I reckon a lot of people are replacing chains prematurely, those tools don't really measure chain 'stretch' they measure roller wear. The only reliable way to measure is with a good ruler. Having said that I just replaced a chain because it was dropping in easily on the 0.75 % side but I got a few thousand winter miles out of it.

Yes, chains don't actually 'stretch' ... it is roller wear that is being measured and those chain neasuring tools work very well and are much simpler and more precise that trying to keep 12 links of a chain straight, in tension, lining up a ruler and trying to take a reading for stretch. However, I would not pay £9.99 for a chain wear tool ... I got mine from On-One for £2 some time back.

If you want your chain to last, then use a good wax dry lube.

 But rolller wear is not what causes 'stretch' It is the accumulation of the wear of the  individual chain bushings that cause the 'stretch' so measuring with a ruler is a less convenient but  much more accurate way of determining when you need a new chain.     

Avatar
lazyusername replied to pockstone | 6 years ago
0 likes

pockstone wrote:

Welsh boy wrote:

BBB wrote:

Rotate 2-4 chains every few hundred miles. The whole system will last longer and you'll never need to measure the chain wear.  

Rotating chains has NO effect on longevity whatsoever, it is yet another one of those myths picked up on clubruns/forums which people pass on as if it were the absolute truth.  Do you really think that taking a chain off and giving it a rest makes it last longer?  Do you really believe that?  Really?

It's Friday tea time , so I'm not in lateral thinking mode, but might rotating chains reduce wear on the chainrings and cassette? 

 

I think if you take the chain off remember which way round it was and remember to turn it round the other way when refitting  it after giving it a good soak, degrease and re lube then yes I'm sure that it may well help to increase longeviety 

 

My personal view is fuck that shit - although I have had to replace a cassette to day so who's laughing now

Avatar
nbrus replied to Anthony.C | 6 years ago
0 likes

Anthony.C wrote:

nbrus wrote:

Anthony.C wrote:

I reckon a lot of people are replacing chains prematurely, those tools don't really measure chain 'stretch' they measure roller wear. The only reliable way to measure is with a good ruler. Having said that I just replaced a chain because it was dropping in easily on the 0.75 % side but I got a few thousand winter miles out of it.

Yes, chains don't actually 'stretch' ... it is roller wear that is being measured and those chain neasuring tools work very well and are much simpler and more precise that trying to keep 12 links of a chain straight, in tension, lining up a ruler and trying to take a reading for stretch. However, I would not pay £9.99 for a chain wear tool ... I got mine from On-One for £2 some time back.

If you want your chain to last, then use a good wax dry lube.

 But rolller wear is not what causes 'stretch' It is the accumulation of the wear of the  individual chain bushings that cause the 'stretch' so measuring with a ruler is a less convenient but  much more accurate way of determining when you need a new chain.     

Both methods measure 'stretch' across several links, so pick whichever one is easiest for you. I wouldn't say one method was better than another, but using a ruler is more reliant on the operator taking a correct reading, whereas a chain guage requires no operator judgment at all ... its simple and repeatable pass/fail, though with the ruler you may be able to measure how much wear has taken place rather than the guage method which only shows when a wear limit is reached. Its the rollers on the chain that mate with the cogs in your gears and that is what a chain guage rests on to check for wear. As long as you have a method that you are happy using, then thats all that matters ... absolute precision isn't really required.

Avatar
Anthony.C | 6 years ago
0 likes

The point is the tools guarantee inaccuracy as they force the rollers apart and roller wear does not affect chain spacing and so shouldn't be included in the measurement.  Only pin/bushing wear affects chain spacing and performance. I use one myself but only consider replacing the chain when it drops in fairly easily on 1% if I can't be bothered measuring properly. 

Avatar
madcarew replied to Anthony.C | 6 years ago
3 likes

Anthony.C wrote:

The point is the tools guarantee inaccuracy as they force the rollers apart and roller wear does not affect chain spacing and so shouldn't be included in the measurement.  Only pin/bushing wear affects chain spacing and performance. I use one myself but only consider replacing the chain when it drops in fairly easily on 1% if I can't be bothered measuring properly. 

But roller / bush wear does affect chain spacing. It is the only thing on your chain that wears. The links don't stretch. If the rollers are worn away inside (or outside) the distance between surfaces that the sprocket or chainwheel teeth catch on are further apart, and so cause wear on the teeth. Both methods, correctly applied will come up with pretty much the same answer. The vast majority of wear is on the inside of the rollers against the bushings and where the pins go through the links as that's where the grease / oil traps the grit and dirt and then the 2 surfaces grind against each other, whereas the interface with the teeth is largely static. 

The 'accuracy' that you talk about is trying to accurately measure an unknown / arbitrary number. We know that chains should be changed at about .75 - 1% wear (33% tolerance in measurement) to reduce drive train wear. This is just a guide number, and if either measuring method is within 10% (i.e. .825 - 1.1%) it still achieves the desired goal of picking a point of wear that is not too extreme but allows a reasonable life for the chain. 

Avatar
madcarew replied to Applecart | 6 years ago
0 likes

Applecart wrote:

Paul__M wrote:

Darkerside wrote:

One of the down sides of a recumbent. Chain swaps need three of the damn things...

Feels a little less disposable at £60+ a pop.

 

That's another reason my High Racer is a 'Sunday' bike. Must get it off the trainer now the clocks have moved.

 

If the chain is 3 times longer, it will last 3 times longer. Think about it  - the average wear on each link is 3 times less as it is in contact with a sprocket under force 3 times less frequently over its lifetime. Therefore the price is the same.

The average wear on each is slightly less, but it is not contact with the sprockets that causes the wear, it is motion under pressure. The length of the chain has an influence, as the point of exit from the sprockets and entry to the chainwheel are the points of highest wear and it occurs less often in a long chain, but chain wear is not proportional to chain length.

Avatar
Jack Osbourne snr | 6 years ago
3 likes

Recently, I deliberately ran a chain for a full year of commuting and after the chain wear indicator hit 1.0, I was able to measure the knock-on effect by the change in shape of chainring teeth. It was dramatic.
Don't underestimate how much damage a worn chain will do to the more expensive bits of your drive train.
Got nearly 4000 miles out of the chain, but the chainrings will need replaced at about 60% of normal lifespan and I got a similarly reduced life from the cassette that was fitted.

For commuting, I'll generally buy whatever is cheapest but really like Ultegra and KMC X10. Currently running a SRAM 1030. Normally, a chain will last me about 6 months/1800 miles with minimal upkeep. I run for 3 months before I start measuring wear.
On the fun bikes, I use Campag and KMC with the Ti-N coating.
ALL my chains, regardless of brand are connected using KMC missing links.

Avatar
nbrus | 6 years ago
0 likes

I have a friend who used old motor oil to lube his chain. He never cleaned his chain and he ran it until it 'looked' worn out (loose and floppy). When he eventually did get around to replacing the chain he found that it skipped on the rear cassette and that only a single link was mating with the front chainring ... he had worn out his drivetrain on a single chain that had lasted only 500 miles (estimated). He has since replaced his worn drivetrain and is now looking after his chain using a good wax dry lube. The bike was an old hybrid, though he does have a race bike that was also treated with neglect, but wasn't used as much, so is still ok. Dirt can quickly grind down your chain and wear out your drivetrain if you let it.

Avatar
Anthony.C replied to madcarew | 6 years ago
0 likes

madcarew wrote:

Anthony.C wrote:

The point is the tools guarantee inaccuracy as they force the rollers apart and roller wear does not affect chain spacing and so shouldn't be included in the measurement.  Only pin/bushing wear affects chain spacing and performance. I use one myself but only consider replacing the chain when it drops in fairly easily on 1% if I can't be bothered measuring properly. 

But roller / bush wear does affect chain spacing. It is the only thing on your chain that wears. The links don't stretch. If the rollers are worn away inside (or outside) the distance between surfaces that the sprocket or chainwheel teeth catch on are further apart, and so cause wear on the teeth. Both methods, correctly applied will come up with pretty much the same answer. The vast majority of wear is on the inside of the rollers against the bushings and where the pins go through the links as that's where the grease / oil traps the grit and dirt and then the 2 surfaces grind against each other, whereas the interface with the teeth is largely static. 

The 'accuracy' that you talk about is trying to accurately measure an unknown / arbitrary number. We know that chains should be changed at about .75 - 1% wear (33% tolerance in measurement) to reduce drive train wear. This is just a guide number, and if either measuring method is within 10% (i.e. .825 - 1.1%) it still achieves the desired goal of picking a point of wear that is not too extreme but allows a reasonable life for the chain. 

This explains better than me why roller/bushing wear does not really affect spacing. http://pardo.net/bike/pic/fail-004/000.html 

Avatar
hawkinspeter replied to Anthony.C | 6 years ago
2 likes

Anthony.C wrote:

This explains better than me why roller/bushing wear does not really affect spacing. http://pardo.net/bike/pic/fail-004/000.html

Thanks for that - a most informative link.

However, I suspect that roller wear is going to happen at more or less the same rate as bushing/pin wear, so I doubt that the margin of error for chain wear tools is going to be that much of a problem. I'm going to carry on using mine as it's a lot easier to use than a ruler.

Avatar
nbrus replied to Anthony.C | 6 years ago
0 likes

Anthony.C wrote:

This explains better than me why roller/bushing wear does not really affect spacing. http://pardo.net/bike/pic/fail-004/000.html 

And from that article...

Using a ruler can be error-prone because it is necessary to hold the ruler precisely and measure one end while making sure the other does not slip. For that reason, several companies have developed chain wear measuring tools. The advantage of a special-purpose tool is that it is faster to measure wear.

Unfortunately, as of 2009/10 only one commercial bike chain wear measures just pin wear. The usual approach is to spread several links of chain by pushing the rollers apart. However, roller wear is added to the measurement, even though it does not affect proper chain operation. The one exception is the Shimano TL-CN40/TL-CN41, which measures only pin wear.

If a tool is conservative, chains are reported as worn-out when they still have useful service life, which increases chain costs. If a tool is not conservative, then errors mean a worn chain can be reported as good, gets used too long, and leads to drivetrain damage.

Commercial tools are conservative and never report a chain as good when it is actually worn. They may, however, report a good chain as worn, leading to increased chain costs.

You may draw your own conclusions.

 

Avatar
madcarew replied to Anthony.C | 6 years ago
2 likes

Anthony.C wrote:

madcarew wrote:

Anthony.C wrote:

The point is the tools guarantee inaccuracy as they force the rollers apart and roller wear does not affect chain spacing and so shouldn't be included in the measurement.  Only pin/bushing wear affects chain spacing and performance. I use one myself but only consider replacing the chain when it drops in fairly easily on 1% if I can't be bothered measuring properly. 

But roller / bush wear does affect chain spacing. It is the only thing on your chain that wears. The links don't stretch. If the rollers are worn away inside (or outside) the distance between surfaces that the sprocket or chainwheel teeth catch on are further apart, and so cause wear on the teeth. Both methods, correctly applied will come up with pretty much the same answer. The vast majority of wear is on the inside of the rollers against the bushings and where the pins go through the links as that's where the grease / oil traps the grit and dirt and then the 2 surfaces grind against each other, whereas the interface with the teeth is largely static. 

The 'accuracy' that you talk about is trying to accurately measure an unknown / arbitrary number. We know that chains should be changed at about .75 - 1% wear (33% tolerance in measurement) to reduce drive train wear. This is just a guide number, and if either measuring method is within 10% (i.e. .825 - 1.1%) it still achieves the desired goal of picking a point of wear that is not too extreme but allows a reasonable life for the chain. 

This explains better than me why roller/bushing wear does not really affect spacing. http://pardo.net/bike/pic/fail-004/000.html 

That's a great technical argument, but is missing the point. It shows (the non effect of ) roller wear compared to a chain without worn bushings. In reality this isn't going to occur. The point is technically valid, but practically is almost without value. The common chain wear tools *seem* to take this in to account, as a 1% worn chain as measured with a standard tool *generally* shows 1% link extension as well. So, in theory Jobst is right (as usual) but in practice the standard tools provide an effective means of determining the chain wear. The point at which you chose to replace the chain is rather arbitrary. I'm a decent Cat 2 racer. I don't change my chains until they are beyond 1% wear. I go through a cluster about every 3-4 years (+/- 30,000 miles) and a chain every 3000 miles. The argument about wear on the rollers vs wear on the chain is rather academic, because as I said, it's chasing an accurate measurement of a poorly defined point.

Avatar
john1967 | 6 years ago
0 likes

had the same chain on me mountain bike for 7 years and use it for commuting every day. Its gets semi regular coatings of house hold 3 in 1 oil and i have absolutely no intention of changing it as i am not a cycling sucker.

Avatar
srchar replied to john1967 | 6 years ago
3 likes

john1967 wrote:

had the same chain on me mountain bike for 7 years and use it for commuting every day. Its gets semi regular coatings of house hold 3 in 1 oil and i have absolutely no intention of changing it as I spend all my time on moneysavingexpert.com

So, you're just going to wait until it snaps (which it will, at some point) before you replace it?  Do let us know whether you merely smack your balls into your headset, or have a proper crash and lose some teeth, when that happens.

Pages

Latest Comments