Even more briefly

Suppose you're a Rockefeller. In other words, money means nothing to you, and a 100K-250K bill is an insignificant expense. You select a bike model and find you are between two frame sizes. You are offered by the factory the option to custom-build a frame. Would you go through the trouble, or will you be able to get a great fit on either the smaller or the larger of the two frames?


Is bike fit akin to bespoke suits? In other words, on the continuum of seat tube lengths (52.0, 52.1, 52.2, 52.3, ... 57.7, 57.8, 57.9, 58.0) there is exactly one size that will fit you just right? Or is there a (significant) range of sizes (say 53 to 56) that could be made to fit you—subject to adjusting the usual parameters (stem, saddle height, etc)?

When I asked the question I was convinced that bike fit is like made-to-measure suits. In an ideal world we'd have tailored bikes. In a less than ideal one we'd not commit to a brand/model, but shop among what's on the market to choose the best fit. I'm now (more or less) convinced that made-to-measure bike frames is an unnecessary luxury. Every cyclist can more or less easily ride two sizes of each make/model (after adjusting the parameters). For some particularly flexible/knowledgeable riders, three sizes might be suitable.

The clarification is triggered by this question.

Long version

When shopping for a bike, do you

  • commit to a make/model from the outset, select a size, then fine tune the bike to fit you, or
  • choose a make/model that makes it possible to pick a size that fits you with minimal adjustments?

The reason it may be sensible to sometimes use the second method is that, perhaps, bikes are manufactured to closely fit one person for size 52 cm , another person for size 54 cm, a third for 56 cm, etc, and a person who is in-between is best served by going to a different brand and model. These might be labeled by tube length or by a letter or two.

bike fit bell curves

In other words, a given size will fit a wide range of people, but the farther one is from the center of the bell curve used in the design of the bike, the more awkward the fit. In particualr, those who lie right between two sizes are best served shopping in a sequence of bell curves where they are closer to the center of the curve. Do you agree?

At an extreme, consider this: suppose that a model is manufactured to very few sizes, say XXS, M, and XXL. It would evidently be the case that many people would be unhappy choosing a size among those three. Now, the graduation is never this extreme, but the same idea applies. Again, do you agree?

  • 3
    Are you aware of Bicycles Chat ? Might be better to discuss such points there. The problem with Q&A format like SE has is that Opinion questions, Shopping questions, and Discussion questions are all considered off-topic, and this one blends a bit of all three together. Whereas in Bicycles Chat its far less structured.
    – Criggie
    Commented Dec 26, 2021 at 23:16
  • 1
    There is more wiggle room in sizing adaptability than you are implying, your hypothesis isn't true. Commented Dec 28, 2021 at 0:48
  • @whatsisname That's an answer right there. In other words you're saying that if a shopper likes a make/model, it is never wise to give up on that make/model just because either of two sizes will require changes. Either size, or both, can be made to fit just right. It is not necessary to shop for another make/model first to find if a closer fit from the outset.
    – Sam7919
    Commented Dec 28, 2021 at 3:55
  • 1
    @sam well its not a straigh-forward yes/no question, or otherwise clearly defined answer. The Bicycles Chat is a valid tool and we generally don't make enough use of it. Can be good for refining a question before posting.
    – Criggie
    Commented Dec 28, 2021 at 19:42
  • @Criggie I still feel the question has a definite answer. It is not at all a chat-type question. Clarified.
    – Sam7919
    Commented Mar 13, 2023 at 15:14

5 Answers 5


Your hypothesis is incorrect. It's especially incorrect on frames with 2cm between different model sizes, but I'd say it's also incorrect on frames with even 4cm between sizes.

Here's why.

Today, bike frames usually have a sloping top tube. It means if you choose a reasonable frame size (not necessarily perfect as perfect may be unavailable due to you being between two sizes), it's completely impossible that the frame would have too little standover clearance. So that's not an issue. Also today seat posts are generally very long, 350mm being common, so it's practically completely impossible that you won't have enough seat post extension to move the saddle up high enough.

So saddle height and standover clearance won't be an issue -- unless you select the frame size to be so far off it would beg the question: why so far off?

Once you have a frame with suitable saddle height and standover clearance, the next thing to adjust is handlebar height. Typically the standard is that there are four 5mm spacers and also the stem is +/-7 degrees. For typical 100mm stem, flipping the stem gives about 25mm extra height. Adjusting the spacers gives 20mm extra height. So you have 45mm range of adjustment for handlebar height. If that's not enough (for example due to the fact that today it's very trendy to have ridiculously low handlebars on drop bar bikes whereas for a rider not especially fit it's better to have handlebars at the saddle level), not all is lost: you can buy a different stem with +/-17 degrees or even +/-35 degrees angle. Also it's almost certain you have to buy a different stem anyway: typically even if you choose the best frame size available, as they are discrete and not continuous, it's almost certain the bottom-bracket-to-handlebar horizontal distance is incorrect. Thus, you have to choose a different length stem after few hundred kilometers of riding once you have decided which length stem is correct. So, you cover two adjustments in one purchase: you purchase a stem with both correct length and correct angle to give you the desired handlebar height and horizontal distance from bottom bracket.

Then the only remaining adjustment is the saddle fore-to-aft adjustment. Generally it's recommended to select bottom-bracket-to-handlebar horizontal distance by whatever feels the most natural when riding standing (climbing or sprinting), and then the saddle fore-to-aft position is determined based on what feels the most natural when sitting, given the bottom-bracket-to-handlebar horizontal distance that was adjusted to be perfect when riding standing.

About the only thing that can go wrong in this saddle fore-to-aft adjustment is that the saddle rails don't have enough adjustment range. If this is the case, you just select a seatpost with different setback.

There! So with two purchases: almost certainly a new stem, and possibly a seatpost with different setback, you made the frame perfect for you. The total cost of these two purchases is less than 100 EUR, and if you only need a stem, you can get one as cheap as 30 EUR.

Comparing these to the price typical quality bikes start their pricing at (about 1500 EUR), it's very cheap.

  • 2
    Wow, finally a post from you I can agree with on all points :) Except maybe the point about handlebars being low on road bikes. On smaller frame sizes the head tube can actually be longer than one would like, but it can easily be solved with a downwards angled stem.
    – Michael
    Commented Dec 26, 2021 at 20:21
  • 2
    @Michael The point is that many cyclists (especially newer ones) strive to have TdF level bar setups because they look cool or whatever. Such a low bar setup is not beneficial (in fact, it is harmful) unless one is thoroughly trained to use it.
    – MaplePanda
    Commented Dec 27, 2021 at 0:19
  • 2
    Quite to the point juhist! A new stem, saddle, and bar tape are nearly always must-buys with a new bike. In particular your advise on riding a few hundred km before fitting it is good.
    – gschenk
    Commented Dec 28, 2021 at 23:15
  • Would you point to a better explanation to the method finding right stem size. I think the brief summary here might confuse readers who need this advise.
    – gschenk
    Commented Dec 28, 2021 at 23:16

I don't believe frames are sized to suit a certain height. On modern bikes, the only measurements of significance regarding fit are the reach and stack (everything else is either a product of these two factors or affects handling, not fit). The height recommendations are really just a guideline for one to follow until they have a good idea of what reach and stack they like to ride.

Furthermore, those height recommendations are just blanket statements. I don't believe any manufacturer is precisely calculating the perfect height range for each size of each model frame they make. Instead, there will just be broad categories (eg. everyone 5'5–5'9 should be on a medium MTB / everyone 5'9–6'0 should be on a size 56 road bike). It's not like "oh, we made the reach 3.14159mm longer on this frame, so the recommended height range should increase by 2.71828mm" or whatever.

Take Specialized for example. They are moving away from size descriptors like "small", "medium", and "large" in favor of just numerical designations (S1, S2, S3...) for each frame dimension combo. If one likes a 440mm reach, then they just find the model closest to that measurement. Sure, if their preference is in between, then a size up vs size down decision will need to be made, but usually that's just a question of 10mm stem change.

Given the example of 175cm, I'd recommend sizing down. Smaller road bikes are generally not a problem, and the absolutely massive MTBs common today even less so.

Lastly, most people are pretty flexible regarding size, so the concept of a "perfect fit" isn't always necessary to achieve. A couple millimeters either way likely isn't a problem, and if it is, a stem swap would fix that issue. The other stuff like cranks and handlebars can similarly be replaced or adjusted.

  • 2
    @Sam If you want to talk about the exact bike in question, feel free to jump into the chat to do so.
    – MaplePanda
    Commented Dec 27, 2021 at 4:44

I've got to go with MaplePanda on this one.

Manufacturers have changed, i.e., evolved, their thinking on sizing. It's starting to resemble putting 10 religious scholars in a locked room, to get 12 opinions in the end.

The real problem here is the difficulty posed for consumers by the non-conformity in bike sizing among manufacturers. Now, it's not just a matter of measuring differently for road bikes vs MTB's (where does that leave gravel bikes?). It's worse than that.

Some use S, M, L, XL. Some use seat tube length. Others use stack and reach. Specialized now has its own proprietary measurement system, S1, S2, etc., which should go a long way (sarcastic) to lessening the confusion. (Hey Specialized - if you're listening - it may be more accurate, but it's not less confusing for buyers).

Try to simply explain all of this sizing nonsense to someone with a revived interest in biking, and who is trying to pick the right size to start with. This should not be a difficult task or explanation. Yet with the current multiple systems, it's nearly impossible.

  • So what’s the solution? As I see it the problem is that you can’t condense a whole set of tube lengths and angles into one number or letter. IMHO stack and reach are the best “universal size” we have. And even stack and reach don’t tell you everything.
    – Michael
    Commented Dec 27, 2021 at 18:48
  • The second best solution would be to define an “average, standard human” in certain heights (1.6m, 1.7m, 1.8m etc.) with certain corresponding limb lengths. We’d then determine certain parameters which have to be fulfilled for a good bike fit (e.g. torso to arm angle should be 90° for a touring bike). The bike manufacturer would then recommend a certain frame size for each standard human to achieve the fit.
    – Michael
    Commented Dec 27, 2021 at 18:51
  • I think the Specialized sizing system is wayyyy better. Ibis is moving in that direction too. I car about: Effective top tube, head angle, reach, wheelbase, and to a lesser extent, chainstay length. Just show me the geo chart and I'll know which too might work. And then I'll pick the one that feels best.
    – Paul H
    Commented Nov 22, 2023 at 16:15

I think this re-statement of the question improves it. The conceptual diagram of how each size could serve a range of riders is useful. And in theory, the further away from the middle point of a size's fit envelope (my term, not an industry term), the less good you might find the fit.

As others have alluded to, however, that diagram (not in the original question, I think) is oversimplified because the fit parameters of interest are multidimensional.

As a worked example, I have a long torso, I am short overall, and I prefer a far forward saddle position. In your conceptual diagram, I am likely to be close to the center of the (presumed) distribution of heights appropriate for small frames. However, because of my long torso, I usually need a 120mm stem, perhaps longer. Bikes in my size would usually come with a 90 or 100mm stem. This slows the handling further. Many performance road bikes have trail in the mid to high 50s for sizes 56 or equivalent, but frames in my size tend to come with trail in the low to mid 60s, sometimes higher. That's already slower handling than stock, which I would be slowing with a longer stem. It is true that most riders can adapt to what they're given, however.

That said, one potential rationale for a custom frame is that the builder - if they are competent and experienced, which may involve working with a good bike fitter - could customize the frame's dimensions to suit you. In my case, I have a custom road bike with about 56m trail with a front end like a medium stock frame, and the rear end of a small triathlon bike. For riders shorter than I, the trade offs inherent in 700c wheels might mean that they choose 650B or 650c. Additionally, I've heard framebuilders assert that most stock frames are over-built for small riders due to ISO requirements, and that they can build frames at the appropriate stiffness. For very tall riders, few manufacturers might make frames large enough, and extremely large riders might benefit from added measures to stiffen the frame (e.g. I once saw a frame with a double top tube). Or, say you are in the middle of the bell curve in many dimensions, but you want some characteristic that's hard to find in stock bikes, e.g. racey stack/reach but stable handling, really steep or really slack seat angle, etc. Before the advent of endurance bikes, people would ask for customs with relatively tall head tubes for a more upright position. People on some forums might say that their bikes would look faster if they went backwards, but those folks were on to something.

Thus, custom frames are probably not necessary for most people. They can benefit more people than that. Also, appreciation for craftsmanship is another reason to get one.


While the height of the seat is adjustable in wide range, the horizontal distance between seat and handlebars can only be adjusted in the range of few cm for seats that can be moved aft and forward, so not all seats. This impacts the mass balance, how much weight do you take on your arms from the seat. You cannot do this by adjusting the height of the seat because it must match your legs. You could do by raising or lowering the handlebars but for some reason they are not always adjustable either.

Hence horizontally adjustable seat must compensate the mismatch of the frame. Of not, then (literally) pain in the ass.

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.