If you haven’t heard, high end mountain bike drivetrains are making the jump from 9 speed, to a full 10. Obviously, the road groups switched over awhile back, with one manufacturer even packing in a whopping 11 gears into one cassette. While most manufacturers will try and convince you that 10 speed is better for mountain bikes, there are quite a few skeptics who still feel that while 10 speed may shift smoother, or faster, it certainly can’t be as durable, right?

Regardless of how you feel towards the matter, the 10 speed revolution is marching forward, but just what does that mean to you?

After having some time on a 10 speed group ourselves, we couldn’t help but wonder how the durability of the chain and cassette really compared to 9 speed. With regular preventative maintenance and cleaning I could get quite a few miles out of a chain, and especially a cassette as long as the chain wasn’t ridden past 1% stretch. Eventually, we decided to put together the best controlled experiment we could, to pit 9 speed and 10 speed Shimano XT parts against each other in a battle to the death.

This is the first post of many, in which we will chronicle the wear and tear of 6 drivetrains, three of which will be 9 speed and the other 3 will be of course 10 speed. We went to great lengths to make sure the test is as fair as possible, in order to get fair, unbiased results.

How is the test going down? Who are the test riders, and what are they riding? Find out after the break!

In order to have the most even and fair testing possible, we sought out 3 pairs of riders, each with similar riding styles and strengths and weaknesses. One pair are excellent climbers, yet somewhat timid descenders, another pair are both endurance specialists with good all around skills, and the final pair (which includes myself) are slow climbers but tenacious on the downhills. In addition to having similar riding styles, we also made sure that one member of each group was riding 9 speed while the other is riding 10 speed.

Equally important as riding style, is what bike everyone is on, as hardtails vs. full suspension bikes can wear chains differently, and even some different full suspension designs can have an altered effect on chain wear. Fortunately, it just so happened to work out that 5 out of the 6 participants in the test are all on 2010 or later, Trek Fuel EX full suspension bikes. The only dark horse of the group is also the only member of the test on 29 inch wheels, and will be riding a 2010 Gary Fisher HiFi Pro. Due to the fact that he will be riding 9 speed, I feel that it won’t alter the results of the test, and will further serve to indicate whether 29’rs wear chains at the same rates as 26 inch bikes.

Finally, all riders involved are all from the same area ensuring that most of their riding will all be done in the same conditions, on the same terrain. While traveling might skew the results somewhat, I did make sure that most of the big events that any one rider is traveling to, the other rider in the pair was also most likely to attend. Considering this is an independent test, we feel that the control over the variables that we have implemented is more than adequate to provide satisfactory results at the end of the test period.

In order to facilitate the controlled test, all riders will be equipped with brand new Shimano XT chains and Cassettes, with the 9 speed riders on 11-34s and the 10 speed riders on 11-36s (testing the same sized cassette is irrelevant due to the difference in front chainrings between the two groups, and Shimano said there should be negligible difference between the two sizes based on gearing) . Since different levels of maintenance can affect chain wear and performance, all 6 of the bikes will be assembled, cleaned, and maintained by me for the duration of the challenge. All bikes will also be lubricated, by me, using only Dumonde Tech Original formula.

Since this is a chain wear test, I will need to keep track of chain wear and to do so, I enlisted the support of FeedBack Sports in the form of their Digital Chain Wear Gauge. The FeedBack Gauge is critical to the test due to the fact that it is much more than a go/no go gauge and will provide accurate measurement of wear down to 0.01mm! This is easily the most accurate chain measurement device on the market, which made it a no brainer to use it for testing.

Clearly, the FeedBack Sports digital chain gauge is a slightly tweaked KMC gauge, although there is a slight advantage to the FeedBack Sports model, and that edge would have to be customer service. With FeedBack Centrally located in Colorado shipping is extremely fast, and if you ever have an issue their team has some of the best service in the industry. Actually, the first gauge that was sent to me, was a fluke from the factory and after a quick email, a brand new gauge was at my door the very next day. Can’t ask for any better service than that!

When using the digital chain gauge, you squeeze the two ends together which sets the measurement at 0, and after placing the gauge on the chain, the spring loaded arm expands pushing the rollers apart and giving you your measurement. In order to ensure even greater accuracy, whenever I measure a chain I will always take at least 4 measurements on different sections of the chain, using the measurement with the most wear. This is due to the fact that some riders tend to have an uneven pedal stroke, which will cause the chain to wear differently at different spots on the chain. Since one worn section of chain will adversely affect your drivetrain, even if most of the chain is still OK, once any section of the chain is worn out the chain will be considered done. As you will see in the rider bios, all of the new chains have been measured to provide their baseline measurement which will allow to see how they are progressing throughout the challenge. FeedBack Sports considers a measurement of 0.80mm to be completely worn out, so that is the magic number that we are looking for to retire each chain.

With all testers set up with their chains and cassettes, the only thing left to do will be to pray for dry trails. Unfortunately our area of the Midwest is home to some of the worst mud I’ve ever tried to ride, so initial test results may be a little slow coming. In regards to keeping track of mileage for the duration of the test, each bike will be fitted with a cycle computer with the wheel size programmed to the actual measurement of the tire, rather than the pre-programmed stock sizes in the computer. At the beginning of the test the odometer will be set to 0, and will not be touched until the end of the challenge.

Finally, our goal is to take measurements of all the chains involved in the test about every other week, or every 50 miles, whichever comes first. When all is said and done, hopefully this will shed some light on whether 10 speed is at least as durable as 9 speed, at least when it comes to Shimano. I think it’s safe to say that we are all just as anxious as you, to see the results.

Stay tuned for the next installment of the BikeRumor Chain Challenge, where we will post the rider bios, bike checks, and starting measurements!


  1. Kovas on

    Commend you on the great idea. I’m anxiously looking forward to these real-world results. I’m guessing that like most, I’m betting 9-speed will be stronger longer, but at the same time… secretly hoping that the new 10-spd is at least on-par with the old… Thanks for running this experiment.

  2. Spencer on

    You really should consider using Dumonde Tech Lite, the yellow formula, on 9-10 speed chains. At least that’s the word from the horse’s mouth. My experience corroborates said horse’s mouth.

  3. Oli on

    Great idea, but I have one question. You say “…some riders tend to have an uneven pedal stroke, which will cause the chain to wear differently at different spots on the chain.”, and my question is that, as the chain rotates at varying speeds to the cranks depending on speed and/or gearing, surely the wear is averaged out over the entire length of the chain? Why would it wear in particular spots? Cheers, Oli

  4. pen1s hurst on

    This makes no sense. Can someone explain the thought behind this?
    I can understand maybe the cassetts being different because there is one extra cog to share turns with (each cog get 1/10th use vs 1/9th in theory), but the chains are only different in width, not different on the parts that wear and cause chain stretch. The side chain plates and inner chain plates are the same. they are just joined with more narrow pins and collars.
    This makes my (head) hurt.

    • Geoffrey Rees on

      Sorry so late coming to this discussion but have just looked at packaging on front of box for Sram PC1051 10spd chain and PC 971. It states on front label bumph that both chains are 114 links, 1/2 inch by 11/128ths. This is sam efigure given by Wipperman for their ‘connex’ chains. It would appear that someone’s taking someone for a ‘ride’?

  5. Devans on

    Look forward to see the result.

    4 measurements per chain will help with inconsistencies in measurement but to say that uneven pedal strokes will wear a chain inconsistently throughout the length of the chain is not accurate. With 27 or 30 speeds, the chain will be operated over a wide range of gear combinations (ratios) and is constantly changing position relative to pedal stroke evening out wear over all the links. Only on a single speed bike where the Chaining teeth count decides evenly into the number of links could this be expected. Your could however see uneven chaining wear due to uneven pedal stroke.

  6. Big f-ing deal!? on

    The guys set out to do a pretty damn well thought out test, and directly a couple of idiots question the testing methods.

    Sure… a lab test in strictly controlled circumstances would give unbiased and scientifically correct results. However, I think most of us ride outside, and not one single rider is exactly like the next.

    Pen1s hurst and dgaddis could do us all a favor and just appreciate that someone takes the time to test it at all. A bit pissed off, but some people just thrive a bit too much on getting caught up on details.

  7. ZachOverholt on

    Not sure of the exact cause, but I have come across many customer’s chains which measure fine in one part of the chain, yet another section is completely worn out. Perhaps there are other forces at work, but I am going off personal experience where it has always been best to take multiple measurements.

  8. ZachOverholt on

    So I just measured a customer’s chain that is ready to be replaced, took 4 measurements with a Digital Chain gauge, and the measurements varied from .50mm to .74mm (.8 is considered completely worn). Again, maybe I’m wrong about the cause of the different points of wear on the chain, but there is a clear difference, and if you only measured once at the point that is .50mm you would (falsely) conclude that the chain is still good, which is why I always take multiple measurements.

  9. Vincent on

    A 29″er whell is taller.
    So to get the same transmission development, one have to use more reduction in the transmission.
    This means that on a 29″er, small chainrings will see more use. This lead to more tension in the chain.

  10. Fabricio on

    Go ahead Zach, even if is a weird test, something interesting will come at the end. It will be good to test the equipment in two exactly bikes, because sometimes the geometry and riding style can influence in the pedaling, also the person (if there are two test drivers) has to be the same physical condition to be close to the performance, aside of the 10 pr 9 speed setup. Well at the end, as I told you good luck and I want to see the results for have ideas or confirm theories.
    Cheers Zach

  11. Devans on

    I am very intrigued that you have seen such a variation in chain wear on one chain. Wonder if inconstant lubrication or sections that get dirtier then other can lead to this. Anyway, good on you for doing the test, I look forward to the results.


  12. ed.rides.bikes on

    I also agree about the chainrings. They are the uncontrolled variable and depending how worn or new they are could make a huge difference in the outcome of this test. Also, are all six drivetrains going to be clean and and lube the same way, with the same lube and in the same manner? How, are you going to factor for different riding (muddy, wet, dry, etc.) conditions?

  13. ZachOverholt on

    Ok, so originally the chain rings weren’t in the budget, but after hearing the comments and thinking about it, we decided to figure out how to make it work. ALL bikes in the test will have completely new rings installed before the start of the test. So that means new cassette, chainrings, and chains, so it should cover all variables. Thanks for the feedback, and be sure to be on the lookout for the next installment!

  14. Brady on

    And of course they all ride at the exact same cadence so the chain goes under the exact same revolutions. Oh and they are in the same gear all the time so the radius of curvature of the chain is the same and the angle the chain is at due to chainline differences.

    Too many variables. This test is going to be useless. If you want true results, you need to do what wipperman did and put gears on the machine: http://www.cantitoeroad.com/news/66/66/Chain-Wear-Test-Results-Point-to-Wippermann.php

    It’s obvious these guys are writers, not scientists.

  15. ZachOverholt on

    @Brady, True, but there is often a difference between Lab tests, and real world experience. There have already been Lab tests that prove Dynasys chains wear better than their 9 speed counterparts in a controlled environment, but how does that translate to the trail? Even though Wipperman’s (own) chainwear tests simulated cross chaining, and outdoor conditions (water, sand, and oil?), the chain still wasn’t actually put through its paces in the real world.
    Sure, there are a few more variables than a perfect scientific controlled test would allow. However, we are trying to control as many variables as possible, and provide some objective (not a company testing their own chains) data for people to make an informed decision. If the results are all over the place, then it will be inconclusive. But, if more than one chain points to similar results, the test will be far from useless.
    And yes, I am not a scientist, thanks for pointing that out.
    Thanks for reading!


COMMENT HERE: (For best results, log in through Wordpress or your social media account. Anonymous/fake email comments may be unapproved or deleted. ALL first-time commenter's posts are held for moderation. Check our Comment Policy for full details.)

This site uses Akismet to reduce spam. Learn how your comment data is processed.