This post will give an overview of bicycle driving chains standard dimensions. Each chain has three important dimensions: pitch, inner width and outer width.
Table Of Contents (T.O.C.):
- Bicycle driving chain pitch
- Chain inner width
- Chain outer width
3.1. Bicycle chain width standards table
1. Bicycle driving chain pitch
Chain pitch is the distance at which the pins are placed. It is measured by measuring the distance between 3 links, then dividing it by two.
Picture 1
For detailed explanation of chain pitch, and why measuring three pins gives a more accurate result read the post Chain wear (“stretching”). For this post it suffices to say that bicycle chain pitch is exactly 1/2 inch (12.7 mm). This goes for all the bicycle chains, regardless of the speed number.
2. Chain inner width
Inner chain width is the spacing between a pair of inner plates. It is marked in the picture 2.
Picture 2
For inner chain width there are the following standard dimensions:
- Single speed chains have inner width of 1/8″ (3.175 mm).
- Multi speed chains, from 5 to 8 have inner width of 3/32″ (2.38 mm).
- Multi speed chains from 9 to 12 speeds have inner width of 11/128″ (2.18 mm).
- “Exotic” standard for freight bicycles is 5/32″ (4 mm).
3. Chain outer width
Chains for one and multiple speeds differ from each other by the outer width. The more “speeds” a chain is designed for, the thinner the outer plates and shorter the pins are (and they protrude less) – so the outer chain width is smaller (i.e. chain is narrower). Inner width of all the multi speed chains is almost the same – with only single speed chains having a significantly larger inner width.
Campagnolo 11 speed, SRAM 10 speed, Shimano 9 sp, SRAM 6/7/8 sp, old 5 speed, 1/8″ single speed chain
Picture 3
Note how rollers of all the multispeed chains are of almost the same width, only single speed chain being significantly wider on the inside.
Pitch is the same for all the chains – they are aligned by length.
As can be seen from the picture 3, the outer width differs mostly. This is important for bikes with multiple sprockets, so the chain doesn’t get stuck (too wide), or drop between the sprockets (too narrow – this latter case is practically never a problem in practice).
3.1. Bicycle chain width standards table
Table 1 gives an overview of chain outer dimensions, by number of speeds.
| Number of sprockets (speeds) a chain is designed for | Chain’s outer width in mm |
|---|---|
| All 6 speed | 7.8 |
| All 7 speed | 7.3 |
| All 8 speed | 7 ~ 7.3 * |
| All 9 speed | 6.5 – 6.7 |
| 10 speed old Campagnolo standard | 6.2 |
| All other 10 speed | 5.84 – 6.1 |
| All 11 speed | 5.46 – 5.74 ** |
| SRAM 12 speed MTB | 5.25 |
| Shimano 12 speed MTB | n/a |
| Campagnolo 12 speed | 5.15 |
| Campagnolo Ekar 1×13 speed | 4.9 |
| Rotor 1×13 | Uses SRAM 12-speed MTB chain standard |
** See the two comments by Klaus here (thanks for the valuable feedback),
until I double check and confirm.
My video explaining bicycle chain construction and dimension standards:
For overview of which chains can be combined with which sprockets, read this post: Bicycle chains compatibility:
I ordered a sprocket using the dimensions provided on this page. I was disappointment when after waiting for it to be milled and shipped the sprocket did not fit my application. When I back tracked I realized because the word inner “diameter” was used. I thought this page was referring to the chain roller. After all diameter is used to describe geometry that is circular and the chain width is not circular. The illustration served to perpetuate the illusion because the lines are coming off the roller. Any ways could you update this page with the roller dimensions.
I have edited the article. In spite of having put “diameter (width)” on two occasions, I agree it is most probably not clear enough. So I have replaced the (confusing, leading to a wrong impression) word “diameter” with the word “width”.
Thank you very much for taking the time to report the problem. I’m sure it will help and prevent anyone else from making a (costly) mistake.
Thank you for the input – any corrections are welcome.
Sorry for your trouble.
I’ll sleep on this and try to understand what was pointing in the wrong direction and how to improve the explanation. Now I still don’t see it – with all the text and images.
Have you measured the roller diameter of a fresh chain?
Width closely matches the inner with of the plates.
Diameter, in terms discussed above – haven’t measured it. Did an experiment though:
https://bike.bikegremlin.com/3370/roller-wear-affect-on-chain-engagement-with-chainrings/
I am attempting to source chain suitable for use with indexing shifters that has a 7.3mm pin length .
Hi, this article got me closer than any other, however, the single thing I wanted to know is missing. I wanted to find out what the outer width of a single speed chain was. Table 3 skipped this entry 🙁 Do you happen to have that chain and could you measure it (not only) for me and fill in the blank in the table?
Thank you very much!
By counting pixels in the picture here I came to a size of 1/3″ (0.3277), that’s 8.3236 mm. Further googling found someone mention it’s about 9 mm. Sounds simillar enough. The maximum width might be even more on the wider join links.
New, KMC single speed is 8.7 mm, while an old one from a kids single speed bike measures at 10.7 mm.
Connect link from the KMC is 9.7 mm wide, so that is effectively the chain’s width for most practical use case considerations.
Single speed chains don’t need to run between (tightly) spaced sprockets so I suppose the outer width is not as crucial.
And it seems to vary quite a bit.
They indeed vary. For example KMC uses indication H for “Heavy Duty” chains and their pin is 9,35mm instead of standard 8,6mm. I suppose other manufacturers have other dimensions. The same goes for 3/32″ chains designed for single-speed use.
Any updates on 12 speed chain dimensions, specifically 12 speed Campagnolo vs KMC 12 speed Missing Link and SRAM 12 speed Power Link?
I just ordered a 12 speed Campagnolo group, and since I need to separate my chain quasi frequently to wax it, I want solution other than riveting Campagnolo pins.
Still haven’t got any reliable data. No 12 speed drivetrains in my shop so far. 🙁
Will have to ask trusted fellow mechanics for measurements and will carry a precise vernier caliper on my next travel abroad and use the occasion and double check.
Only Shimano 12 speed is available in some shops here – didn’t see SRAM, or Campagnolo – Campagnolo is sort of a cycling unicorn in Serbia: everyone knows it’s cool and nice, but (almost) no one has seen it. 🙂
For the purpose of Narrow Wide drivetrains, should more consideration to the width between the inside edges of the “outer plates” be noted? Most Narrow Wide Chainrings state compatibility with 9-11 or 12 speeds. I assume that the “wide” (which I take to mean “thick”) tooth profile fits inside of the “outer plates.” Perhaps these chainrings are not compatible with 8 speed chains because the rings do not fully fill the gap in such chains. Less likely, the “narrow” tooth of such a NW chainring may be 11/128″ thick instead of 1/8″ again not filling the gap between (this time between the inner plates) even if those don’t vary by much. Failure to fully fill either gap may result in failure to retain the chain, thus incompatibility.
Correcting error: meant to say 3/32″ where I wrote 1/8″. Apologies.
It makes sense – narrow wide chainrings only come in even tooth counts (no odd ones), for exactly the reason that it’s the only way to make sure that wide teeth always go between the wider (outer) chain plates, and narrow teeth between the narrower (inner) chain plates.
I would expect 8 (and fewer) speed chains (with 3/32″ i.e. 2.38 mm inner width) to be a less tight fit on those chainring teeth, designed for 9+ speed chains (with 1/128″ i.e. 2.18 mm inner width).
Unfortunately I haven’t had much experience with the modern 1x drivetrains. Probably will in a few years, as more and more people get that stuff 2nd hand from Germany and the likes. The very thing that made me experiment and document compatibility – a constant lack of “proper” components availability – is what makes it difficult for me to test the newest stuff as it comes out to the market. And the fact I make sure to note all the downsides, and call out any marketing gimmicks I notice, makes me believe that it’s not very likely for the manufacturers to be sending me samples for testing, as they do with some other websites/companies. 🙂
So I haven’t made any measurements, or experiments with 1x drivetrains yet. I have found an article that does show a few types and different 1x solutions, though without considering compatibility at all, but it’s a start:
https://www.garbaruk.com/blog/articles-1/post/the-state-of-narrow-wide-chainrings-1
Relja, many thanks for the careful, meticulous and informative work you do in helping us understand the somewhat obscure details of the bicycle industry. in the spirit of constructively and cooperatively helpful assistance, i refer to your Update of 05 Jan ’21, where you note, “Each chain has three important dimensions: pitch, inner diameter and outer diameter.”
First let me say that your command of my first (and only) language is way, way better than any grasp i might have of what is probably *your* first language.
then, i think when you say ‘diameter’ here, you are referring to ‘width’. in english when we say *diameter*, we are referring to the distance across a circle, on a straight line across the center. so unless you are referring to the inner and outer diameter of the chain *roller*, which is circular, the word ‘width’ makes more sense in this context. just trying to be helpful. again, thanks for your public-spirited work.
Hello Chris,
Thank you very much for noticing the mistake and taking the time to let me know. I’ve edited the text right away.
The mistake you noted is not a simple, obvious, spelling one, hence it is more “dangerous” – for it could make the reader draw a wrong conclusion.
I do 99.9 % of both writing and editing myself, so the support and corrections I got from the community over the years have been of huge help.
glad to have been of assistance.
i must say, reading your analysis has led me to change my thinking about multi-speed chains & cassettes. there is a persistent folklore in the bike world that 12, 11, and 10-speed systems have a shorter component life (chains & cogs, when operating conditions are identical) than 5, 6 & 7-speed systems (for example), due to presumably smaller surface areas of contact between teeth & chain rollers (narrower, we assume, when there are more cogs). however, if it is in fact true that roller width (& therefore cog width) is the same for all systems, then there is no reduction in durability of systems with higher numbers of speeds?
The short answer is:
I don’t know.
Longer one:
I’m not 100% sure.
Even longer, with an explanation:
When doing experiments and comparisons, it is very difficult to draw a correct conclusion when more than one thing is altered.
With bicycle drivetrains:
1) Chains
5 to 8 speed chain inner width is narrower than it is for single speed chains.
9+ speed chains are a bit further narrower.
2) Cassettes
Cassette sprockets are practically narrower for each extra gear added (two adjacent numbers of gears sometimes have the same sprocket thickness, like Shimano 10, and 11-speed cassette sprockets both being 1.6 mm thick).
3) Drivetrains
Modern drivetrains often have only 2 front chainrings, or even only 1, and cross chaining is more exaggerated with the modern systems (50-34 compact doubles, or 1x systems). Higher amount of cross-chaining would affect drivetrain durability even if all else were equal – but it’s not, depending on the number of gears, gearing configuration etc.
4) Manufacturing quality
I’ve run comparison tests with 8 speed chains. Primitive, no lab. Just using one same bicycle (mine), same drivetrain (tooth count, cassette and cranks model etc.), same maintenance schedule, and the same riding conditions as much as that can be achieved without a lab. There’s a huge difference in chain life with different models, even from the same manufacturer.
I can’t test 1000 chains, to have statistically valid and reliable data, so I take my own results with a grain of salt. But, when I test the same model twice, and it lasts twice as long compared to another model I’ve also tested only twice, I do believe it is correct to conclude that it is longer lasting. Perhaps not double, not exactly, but to roughly conclude it is better is a conclusion I’m satisfied with.
I have also put a high-end 10-speed chain for a test, on that same drivetrain. To compare how long it lasts compared with the previously tested 8 speed chains, using the same cassette sprocket and crank chainring thickness. If health and luck serve me well, I’ll have results by the end of this year (almost got run over this Friday, on the way to work – careless driver – so I shouldn’t underestimate luck! 🙂 ).
After testing the 10-speed chain on a 7-speed cassette, I plan to re-do the test using a 10 speed cassette (friction shifters make playing with different setups very simple 🙂 ).
Thanks to that, I’ll also be able to test an 8-speed chain on 10 speed cassette sprockets (just running fewer of those, and using some wider spacers).
5) Chain nemesis
My theory is that dirt is the nemesis of chains. All the modern chains, from 7 to 12 speeds have no mechanism of keeping the lubricant inside, and dirt out. The bushings were cast out for lower weight, cheaper manufacturing, and easier severe cross-chaining.
On that account, on my motorcycle, cam timing chain (is that the right term – the chain inside the engine block) is half as thick compared to the drive chain, yet it lasts about 10 times longer. After having installed Scottoiler chain lubricant – that drips low viscosity lubricant over the chain practically constantly, and that swings off the chain along with the dirt, my drive chain lifetime had gone up by at least 3 times, compared to cleaning and lubing it with thicker lubricant on a regular basis.
Doing something like that with bicycles is not very practical – for me at least. So I choose to just run with the cheaper stuff. Decent quality chains seem to last approximately the same, regardless of the number of speeds. However, that’s just my subjective impression, so I’ve decided to run a sort of a primitive test to see if that’s at least roughly correct.
The article was great. Can you shed some light on how to clean and lube a chain. On a single speed bike would a master link be better. so you could take the chain off to clean and lube it. Thanks vince
Hi Vince,
I already wrote in great detail about various different chain cleaning methods, their pros and cons.
You can read it and decide for yourself, what works best for you:
Bicycle chain cleaning and lubricating.
Relja
relja, i just got around to reading your comments of 04 Apr, today, and found them quite enlightening. i agree there are many practical reasons why a truly rigorous, scientific study of chain durability is a considerable challenge for the individual investigator. even so , you have given us a lot to think about, and i am still thinking about it. i’m even thinking of making a hardcopy, to carry around with me and ponder (i don’t carry the kind of ‘device’ that would readily display this much text).
regarding your motorcycle cam timing chain (yes, you had the right nomenclature, we usually just shorten it to ‘timing chain’), it’s no surprise at all that they outlast the external drive chain by a considerable factor. after all, it’s cosily protected from dirt & moisture, and the copious flow of engine oil both lubes continuously, and carries away worn-off particles of its own, abrasive, metal, we hope to be trapped in the engine oil filter.
the great, now deceased, American authority on all things bicycle, sheldon brown, stated that his ideal bicycle drive chain would be housed in a closed oil bath, protected from the elements, and would therefore need to drive an internally geared hub (not practical with a derailleur).
i’d be very interested to see a summary of your results from your trials of various chain makes & models, regarding the (understandably imprecise) relative durabilities. and maybe an index of cost per kilometer? i’ve often wondered if the amazingly costly stainless Wippermann might actually have a lower cost per distance. ??
and thanks again for all your wisdom. stay healthy
where to find roller diameter or do you know for 9 speed chain??
Hi Aditya,
I haven’t measured that – and don’t know if the info is available elsewhere.
My current schedule is hectic, can’t promise any deadlines, but I could measure it, of course, and publish the info.
Relja
a chain is a chain,i dont think any expensive chains are worth the extra money the reason being its still going to wear at the same rate as a cheaper chain,you just go buy the cheapest chain that fits your bike,and you buy a chain that is easy to find quick links for,for me i have never had a problem using even the cheapest shimano chains,infact the cheaper chains are much easier to work with
bicycle chains are measured by chain pin width,example a shimano hg 7-8 speed chain is 7.3 mm wide,more speeds you have the thinner the chain becomes,less speeds the wider the chain,when buying chains allways buy the longest chain and shorten it.
May I ask how much do each size of the chains cost on average?
Hi Michael,
I’d have to put in some time and effort to check the current prices, across brands and models in order to calculate.
Roughly, generally, same brand, same model tier, will cost more as the number of speeds increases.
Taking KMC X-series “Silver” chains for example:
– 7/8-speeds: 15 euros
– 9 speeds: 23 euros
– 10 speeds: 27 euros
– 11 speeds: 32 euros
– 12 speeds: 38 euros
The noted X-series silver chains are of decent quality and their price is what I consider to be reasonable for the quality and durability.
In my experience, the situation is similar with other manufacturers.
* The exact prices may vary depending on many different factors – including taxes and local regulations.
I’m look in ng for a 2″ gear that will roll on a bike chain but only a 1/2″ slit X 1/16″ on the id instead of a circular hole.. can anyone help?
Hi, I heard that about 3 years ago SRAM increased the roller’s diameter of their 11 speed chain from 7.64mm to 7.7mm, and hence you cannot check the wear with a simple 2 point wear check tool, you need 3 point tool. Is this something you also noticed?
Hi Leon,
That’s a good question.
I don’t understand why manufacturers keep selling the inherently imprecise chain wear tool models. In a separate article, I’ve explained how to accurately check chain wear (and why the common 2-point checking tools are bad).
I either use the proper, Shimano’s chain checker, or a steel ruler. Common chain checkers are used just to let me know if I should double-check with a proper tool.
I’ve also made an experiment to see if roller wear affects chain engagement.
Relja
The other day I got a new M9100 12-speed Shimano chain and had to remove a few links. Much to my surprise and disappointment, my Park CT 4.3 doesn’t fit properly. It got the job done, but not in a great way. A bit of a kludge. I later found on the Park Tool site that the CT 4.3 is incompatible and has been discontinued. I guess with the various new chains on the market, the shapes have changed and by having a tool that cast in the shape of a symmetrical link, they’ve made the tool obsolete. I’m thinking of trying to modify my CT 4.3 to make it still useful. Have you ever compared the side profile of a 12 & 11 speed link and compared where they differ? Thanks
Hi Ric,
Here, I wrote about the Shimano XTR CN-M9100 chains.
So, it’s not about the “number of speeds” as the chain is designed to work with both 11 and 12-speed cassettes. The “catch” is most probably in the plate profile.
In the mean time, I bought Unior Master Chain Tool 1647/2BBI (link to my video presentation). However, I’m yet to test it with an M9100 chain, so can’t confirm if it works smoothly (but it does work with SRAM flat-top and Campagnolo chains at least). As those chains are very rare in my country, I may need to buy one just to test the tool (and send the info back to Unior 🙂 ).
Relja
Thanks very much. Im going to try to compare the plate profile and see if there’s anything I can do to file out mine