New XTR groups come once every four years, and as soon as one launches, they start working on the next one. It starts with concept work, thinking about what the key elements need to be well before they get into the technical details. And they need to think ahead to what the bikes will be like four years from now, what racing will be like and how people will be riding them.

XTR will always be about riding faster. It will always have a focus on being the lightest, most efficient system it can be. But with enduro blowing up, they had to consider that use case. With XTR, they also want it to look good. So there’s potentially competing priorities – design, functionality, and light weight. At the end of the day, they’re willing to forgo some non-functional aesthetics to save weight. After all, it can’t gain weight over a prior generation.

2019 Shimano XTR M9100 development story and tech

Photo courtesy of Shimano.

While some riders think Shimano operates in their own little world, they really do look at what the other component companies are doing to make sure their designs won’t be so far off the norm that they appear out of place. Whether it’s suspension or frames or other things you hang on your bike, those are considered when drawing the lines for XTR.

How Shimano develops new products

2019 Shimano XTR M9100 development story and tech

A fully machined late stage XTR M9100 prototype rear derailleur. Photo courtesy of Shimano.

With so many new features, we had to ask: What comes first, the design or the capabilities? It’s both, actually. With the new XTR, research and development started both with what they wanted out of the group, and with new manufacturing tricks they’d learned. The engineering team had developed new capabilities and processes that allowed them to do things they’d never done before, and they wanted to use them. The design and marketing teams knew it had to improve upon the looks and meet modern riders’ expectations.

2019 Shimano XTR M9100 development story and tech

Photo courtesy of Shimano.

Initially, both sides dream up new designs, uses and features. Then they come together to take the look and feel of the design crew and marry it with the tech features in a way that can be mass produced. Not every feature or design cue makes the cut -some simply aren’t feasible at production scale- but both sides are fairly represented in a final product.

2019 Shimano XTR M9100 development story and tech

Photo courtesy of Shimano.

First, they sketch things out on paper, then bring that into the computer and make rapid prototypes to see how things look in 3D. From there, they progress like any other brand, machining prototypes and testing them to refine the mechanisms, pivot points, and functionality. Slowly but surely, all of these pieces come together to finish the puzzle. Here’s how that looks on the bike…

XTR M9100 on the bike

We covered the technical features in our launch story. Here, we’ll add a bit of commentary on how all of the parts come together on the bike. The cranks retain their spindle-bonded-to-driveside layout, with the non-drive arm attaching to it. The biggest two differences are that they’re now using a direct-mount chainring, and a more traditional bolt-on arm design that relies on an adjustment ring rather than their plastic attachment bolt and pinch bolt design. All are welcome improvements and updates.

That does mean you’ll have to pull the driveside all the way out of the BB in order to change the ring, and our guess is you’ll want a special tool to do that. The lock ring is a common BB tool interface, but it’s extremely thin, and tucked in there kinda tight.

One of the key features of the new group is the chain. It’s specific to this group (although word on the streets is it’ll work just fine with Race Face 1x narrow-wide chainrings, too) and is designed to better capture the wider teeth. Click to enlarge the photo above and you’ll see how the big teeth are centered inside the outer plates…and really aren’t even making contact with them. Instead, chamfers on the inner chain plates rest against the leading and trailing faces of the wider teeth, helping the chain center itself and avoid vibrations and noise.

The cassette’s most popular option is sure to be the 10-51. Partly because that’s the range most of us are used to now, and partly because FOMO. None of wants to be caught out there without enough gears, and few of us are racers at the highest level where the exact optimized spread is going to make or break our days. So, the gear steps we’re concerned with are 10-12-14-16-18-21-24-28-33-39-45-51.

Yes, those are some big jumps as you get toward the end of the cassette, but from a percentage increase standpoint, they’re within Shimano’s cadence optimized Rhythm Step program. Starting at the 10-tooth cog, you’re looking at increases of 2-2-2-2-3-3-4-5-6-6-6. Which is technically their explanation for going to 51 teeth instead of 50.

The rear derailleurs use 13-tooth pulley wheels to add more chain length without having to further elongate the cage.

The clutch mechanism is easily adjustable thanks to a port on the bottom, which has a rubber seal to keep crud out.

Cable management is simple, which should make installation a breeze. Adjustment screws are all easily accessible, too.

The new brakes use much of the same master cylinder tech, but move the clamps inboard quite a bit. This not only provides more room next to your grip, but also allows the lever body to rest against the bar for better support:

Shimano says some racers were complaining of too much flex when braking hard, so this solution let’s them keep a slim mounting interface while improving stiffness.

The calipers get a new design that moves hose routing inboard for a cleaner appearance. Bleed ports are easily accessible and top facing.

The XC brakes also lose the banjo for a direct mount hose, but the 4-piston enduro brakes keep the adjustable angle banjo fitting.

Both the XC pedals and Trail/Enduro pedals are heavily updated in appearance, with more pronounced platforms. And they get more spindle options for the XC set, letting racers get the narrowest possible Q-factor.

XTR M9100 actual weights & pricing

First things first – all of these parts weighed are technically prototypes, so final production weights may vary. But, this should give you a good idea of how they’ll compare. Official claimed weights are shown on the price sheet at bottom. Cranks, complete with a 32-tooth chainring, came in at 511g.

The 10-51 12-speed cassette is 316g 376g, the new chain is 262g, and their high-mount chain guide is 28g. There are other options for many of these components, but only select versions were on hand for inspection. We weighed everything we could.

Low direct mount front derailleur is 86g, and the rear derailleur with SGS long car is 243g.

From left to right: Rear shifter iSpec mount (111g); Rear shifter clamp mount (116g); Front shifter iSpec mount (68g); and Dropper remote lever iSpec mount (35g).

A front 4-piston XTR enduro brake came in at 244g (no rotors were loose and available for weighing), and the XC Race pedals came in at 151g per side.

The new Scylence rear hub with standard J-bend spoke flanges was 223g, and front was 131g. Both are with thru axles, and I think both of these are the non-Boost spacing, so add a couple grams for Boost hubs. Considering the tech inside, these weights are impressively low and should provide plenty of reason to consider these if you’re building up a set of custom wheels.

So, how much does the new XTR M9100 cost?

2019 Shimano XTR M9100 pricing and official claimed weights

Add it all up and you have the following for a rideable group without hubs or extras like pedals, chain guides, etc:

  • 1x XC build – $2,015
  • 1x Enduro build – $2,075

Add $310 for a 2x system (switches to 2x front crankset and adds a left shifter). Look for our first ride impressions in June, and leave any questions in the comments and we’ll be sure to get them answered at the ride event. In the meantime, check out the official launch coverage here, and our factory tour from Osaka, Japan here!



  1. VeloKitty on

    Thumbs down. A new freehub body style… what a turd.
    Is the spacing of the cassette the same as SRAM and Campagnolo 12-speed?

    Why can’t the major manufactures work together on standards?
    * bottom bracket standards
    * freehub body standards
    * cassette standards
    * shift lever pull ratio standards

    Electronic companies produce interchangeable parts for PC’s. Electronic standards like PCI, USB, etc. are thousands!!! of times more complicated than the bike realm.

    15 years from now, good luck finding parts.

    • JNH on

      The last time Shimano changed it’s MTB freehub was the advent of 8 speed in 1993, when we all bought AM386 computers. Find me a single PC component from that year that can be fitted to 2018 PC without the aid of an adapter if it’s possible at all.
      In that time just the screen connectors have gone from VGA to DB13, DIN4pin, DFP, DVI, VESA 35 pin, to dual link DVI, Mini DVI, Display Port, HDBaseT, HDMI, Mini DP/DP2.0, HDMI 1.2/3/4/2.0/1, USB-C3.1 and Thunderbolt. That’s ignoring Apple’s proprietary display connectors. USB is the one relatively stable connector standard, but even that equates to over a dozen port standards now and they are not all cross compatible. If you want to talk about PCI, what about PCI-E? Or AGP, ISA, VLB, EISA and MCA? Or what about software based issues that mean hardware compatible devices cannot actually talk to each other?
      The tech industry makes bike companies look like rank amateurs when it comes obsolescence and cross compatibility issues. Shimano having a new freehub standard is a bit of a pain, but when SRAM invent a new standard seemingly with every product release I can live with it.

    • D-con on

      For what it’s worth, it is Sam’s controlling the IP around XD cassettes that prevents Shimano from using the standard. Anyone can make XD hubs, but good luck working around their parents to make anything that attaches to them.

      People (SRAM included) have been using the Hyperglide spline for decades.

        • thestansmonster on

          lol, a failed standard that is now being used by every wheel and hub manufacturer, can be found on MTB, road, gravel, and cross bikes, and is on over half of the bikes at the XC World Cup level. Total failure…

          • Razorback on

            Failed standard because it is a non reliable standard. XD cassettes are expensive to produce, simply because it is a poor design. Have seen enough failed lockring on XD cassettes to call it a poor design.

            What about the outer bearings not lasting enough on XD freehub body?

            A new freehub standard from Shimano every 25 years I’m fine with that.

    • Kevin Kitura on

      I am going to to go out on a limb and say it might have something to do with where the factory and the manufactering engineers live and work.

  2. Loki on

    I don’t think that’s an entirely fair comparison. Although USB is run by a body focusing on standardizing protocols and connections we still have USB 1.x, USB 2.0, USB 3.x, mini USB, micro USB, USB-C, USB on the go, A versions, B versions, AB versions etc. (Let alone lightning connectors, Firewire, Midi, Thunderbolt, eSATA, eSATAp etc.)

    Those variations all come about as more is asked of the equipment and the current protocols cannot handle the new demands. If we we’re all happy to ride around with 9 speed we could keep the existing free hub body design.

    Who decides why we need the new capabilities is a separate debate.

  3. Kevin Kitura on

    Technology marches on! The Shimano Hyper Glide cassette hub standard has been in desperate need of an engineering update for many years now, thank goodness it finally got it.

    We aslo need to keep in mind that when Shimano makes a change it is ussally for engineering and manufactering reasons not for cosmetic and marketing reasons.

    • Jones MacArthur on

      Desperate need? Really? I’ve been fixing bikes for 14 years, a large portion of the life of the current 8/9/10/11/12(if you use a Sunrace cassette) freehub, and the only problem is scarring, which is easily remedied by using steel or ti for the body, or making carrier-based cassettes. The HG freehub works, and that’s all it needs to do.

      • Greg on

        The scarring is a real problem that should be solved with better engineering, not simply a different material. Shimano tried to modify their freehub when they went 10speed. Remember road 10speed-only freehub bodies? They had taller splines…
        Aluminum is a better freehub material in every way, as long as the spline shape is suitable.

        • Antoine on

          Most of the time scarring comes from too low torquing. Also some hub manufacturers simply add one steel spline to protect the aluminium ones, works well too.

          • i on

            and Shimano designed something better. The freehub needed to change to make room for a 10t cog, which a lot of people want (regardless if you think the demand is justified, it exists). So they took the opportunity to improve, rather than the poorly thought-out xd driver.

            And seriously why all the whining? So you can’t use the 20 year old cassette you still have lying around? By the time these are available, I bet every hub that has an xd option will have a option for the new shimano driver.

        • Razorback on


          The HG freehubs made by Shimano were always made out of Steel, not aluminium. If some freehubs have scarring issues, it is because some companies like Hope, DT, etc … made their freehubs out of aluminium..

          HG cassettes were designed with a steel made fhb.

  4. JNH on

    The Chainring lockring tool will apparently be included in the box with aftermarket cranks and fits onto the lockring better than a normal Hollowtech wrench or BB tool would. The lockring is steel too, so it should be pretty difficult to damage, although I’m sure someone somewhere will manage it.

  5. Antoine on

    I have a hard time understanding the 2-2-2-2-3-3-4-5-6-6-6 logic. Obviously the first “2” 10 to 12 is quite big then its quite closely spaced but the first 6 is a bit big and last 6 is a bit small. I think shimano reasoning around having some close spaces for climbing steep is a bit wrong because when terrain is steep and speed is slow bike speed never settle so cadence is not stable at all. Trying to keep it in a narrow window is wasting gears when 10 to 12 can kill a sprint for example. Twelve overcadence then 10 will kill you. Then not everyone is a racer but some are.

    • Marc L on

      I think that the steps are better presented in terms of percentage: 20/17/14/13/17/14/17/18/18/15/13- so you’re you’re right: the 10-12 is big(ish) -but closer to the mean than a 1t jump would be- and the 45-51 small(ish).

      Shimano needs to balance the desire for consistent steps (an average of 16%) with the need for each cog to be a whole number of teeth and working the clocking and ramps of adjacent cogs to allow for smooth single and multiple shifts. You could make the steps more consistent by making the tooth pitch smaller (20-102 cassettes, anyone?) but that has its own set of technical and rider acceptance challenges.

      There are a lot of competing constraints- and given how well Shimano stuff tends to shift I’m inclined to give them the benefit of the doubt.

    • John on

      My guess is that, for a lot of MTB riders, the 20% jump from 12 to 10 will be in a spun out section of trail where the ~13% change from 46 to 51 will be useful for seated climbs up steep pitches.

      If Shimano brings this as 1x with Dura-Ace later this fall for gravel racing, the same logic applies.

      Anywhere that a 11 to 10 jump would be useful would be more appropriate for road sprinters who will very likely be running a front derailleur anyway.

  6. Antoine on

    I also hope they’ll sort some solution for their hub system. It looks fine and all but for someone who finally bought its boost pacing racing wheelset will simply stay sram for not changing wheelset. Maybe spacing is the same as eagle ? Or they better have to make deals with many more companies than DT for freehub.

  7. phillipfogg on

    Whether Sram, Shimano or whatever, they all shift and they shift well. It boils down to personal preference. I think the new XTR group looks awesome! The price not so, so I’ll wait a couple of years till it “trickles” down to the XT and or SLX groups.

    • jxjjd on

      yea the cringy part is”its better cuz i like brand x” from both camps. makes ppl look stupid.
      id be surprised if this groupset is bad. i suspect its quite similar to srams in actual use

    • Bill on

      Honestly the one thing I still use Shimano over sram for is I like the feel on the shifters better. It’s all about being able to use my trigger finger to upshift,I don’t like on my SRAM bikes how the position of the up shift lever takes me moving my thumb so far under the shifter. I want it to be super-easy to drop the cog down in the back when I crest a hill.

  8. CoffeeAddikkt on

    the biggest issue with this whole groupset is the new Crank arms surely!? the only system that has limisted issues in my experience has been the shimano plastic bolt pre-load and cinch system, a little spinning pre-loader that is found in the sram system is forever causing headaches!

    • Crash Bandicoot on

      Agreed. I really hope we don’t see this as a trend with Shimano hollowtech II Is a fantastic standard I can use one torque wrench on 95% of my bike and have had zero issues with it in 40,000 miles of use.

  9. Hpbiker on

    Prices will be 30% less online. Shimano does not protect the value of their products. They do care. The group does look very nice. These look to be some great updates.

    • dirty sanchez on

      Won’t be a long wait, you’ll probably be able to get it 60% off online before us bike shop people will even be allowed to order it.

  10. Kovas on

    This XTR checks a lot of boxes for me, unfortunately also checks the “too late” box…
    I’m guessing my story is like many: I waited about a year before upgrading to Eagle, watching the prices settle and see if the group needed any tweaking. Now a few seasons riding it, I absolutely love it. It’s a hard sell for XTR to come in years later and say, this is better and worth the “upgrade” when many of us are already flying XX1 without complaint…

  11. Me on

    A front derailleur? Really? A derailleur for the front of the bike? WTF? Why bother? Otherwise, it does look as though Shimano is slowly but surely catching up to SRAM.

  12. badbikemechanicx on

    I will likely buy this system. Remember when granny gears were for #$%Ys? I am not sure the slowing of mountain bike riding is a good thing.


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.