I get a weird sensation on the left pedal when I'm riding one of my bikes, as if the pedal was not perpendicular to the crank arm. It feels like the pedal is moving under my foot, although there is no play anywhere, and nothing is loose. Through the pedalling motion, at one point I feel slightly more pressure on the outside of the shoe, then as the crank arm rotates the pressure shifts to the inside of the shoe.

It is a subtle feeling, nothing crazy, but I notice it and it annoys me. It's most noticeable when I change bikes. After I ride another ("non buggy") bike for a while, and then switch to this "buggy" one, I notice immediately the weird sensation in the left pedal (right one too, but less pronounced). Then I forget about it after 5 minutes of riding. I get used to the "buggy" bike. Or maybe both bikes are "buggy", just in slightly different ways.

I wonder if I can measure somehow, with decent precision, the alignment of the pedal in relation to the bottom bracket. The pedals' spindles should be parallel to the bottom bracket axle.

Alignment drawing

I have tried to do a measurement like this:

  1. have the bike fixed on the hometrainer
  2. place a spirit level along the pedal axle
  3. keep the spirit level on top of the pedal (like your shoe when riding)
  4. see if the bubble stays in the same place as the crank rotates

If the bubble stays in the same place then the pedal is parallel to the bottom bracket (right?).

Here is my result showing the maximum range of bubble movement as I rotate the crank:

I am not sure what conclusion to draw from this measurement. If I understand it right, it shows that the two rotating things (1. crank, 2. pedal) have rotation axes that are not parallel to each other. If they were parallel, then the bubble should have stayed in the same place (not necessarily in the middle, but same place). Is that correct? What other measurements can I do to check this?

Fun fact: none of my four bikes pass this test. Each one has some bubble movement, but the "buggy" one has a bit more bubble movement. Are these just normal manufacturing tolerances? Are all my bikes just misaligned? Is my test flawed?

  • 1
    If you have 4 bikes, a good additional test would be to swap pedals and see if you see the same error. That would allow to check if the origin is the pedal or the crank.
    – Rеnаud
    Oct 27, 2023 at 8:39
  • 1
    Also check for damage. On my roadbike I had the feeling that the right pedal was wonky at the bottom part of the stroke: turns out it was a small crack in the now recalled Shimano crank (you don't seem to have an affected model). It is possible a tiny crack would only be visible when pressure is applied.
    – linac
    Oct 27, 2023 at 11:25
  • 1
    The bike needs to be vertical for your pedal test to work. Use the level to verify that the seat tube is vertical and then re-measure the pedals. I suspect Andy is correct, the pedal spindle/axle is bent.
    – David D
    Oct 27, 2023 at 12:41
  • Would not the error be constant in one direction if the pedal axle was bent? I checked my cheap bike with a phone-level app, no deviations on either L/R side.
    – WornChain
    Oct 28, 2023 at 8:29

4 Answers 4


The big issue with the test you're doing is that misalignment of this type, resulting in the warbly pedal feeling, can be either in the pedal spindle, the crank, or the BB spindle, and if all you're doing is putting a level on top of the pedal your reading is going to combine the net effect of all three and not give you any ability to isolate the problem so you know what to replace.

Switch the pedal with a known good one and ride it to test the pedal.

Remove the crank bolt and put in a sufficiently long M8x1 bolt or threaded rod, then see if it warbles up and down as the cranks spin to test the spindle. (Don't ride the bike with the crank bolt removed, just spin the cranks backwards with the bike stationary).

Use process of elimination with the above two steps to determine if the distortion is in the crank arm. You could rig something to measure the alignment, i.e. a BB spindle fixed to some kind of reference surface, but there's really no need.

It's possible it's more than one thing, which would usually go along with the crank or pedal having taken an impact. If that's true you're still following the same logic as above until it feels normal when you ride.

Crossthreaded/under-torqued pedals that result in this issue became a much more common thing to see in the post-pandemic years, with such a higher percent of bikes sold direct to consumer. If the problem is with the NDS crank pedal threads, it's possible to repair them via helical insert, but the more economical way is usually just replace the crank arm.


The situation in the center diagram can arise if the pedal spindle is bent. This can occur as a result of crashing. Or (a riding partner experienced this) damage inflicted by airline baggage handlers. You could remove the pedal from the crank, hold it in one hand and spin the pedal spindle. If the bend is bad enough, you might see that the end of the spindle does not rotate in line with the expected spindle centerline.

My riding partner developed knee pain as a result of the bent pedal. This cleared up when the pedals were replaced with a new, undamaged pair.

  • It can occur as a result of crashing. On less expensive bikes or on less expensive pedals I've seen new pedals with bent spindles brand new. Off the bike you can spin the pedal spindle in your hand and see that it's bent.
    – David D
    Oct 27, 2023 at 12:43

The question does also relate to frame alignment. I don't have a comprehensive answer, but we can discuss some issues.

I think you're right that if everything is aligned, the spirit level bubble should be dead center through the full pedal rotation. But that means the frame has to be vertical as well.

Without meaning to gaslight you, the sensations that we feel can be misinterpreted. Remember that we are sitting on the bike and trying to judge sensation from things that our nervous systems are not connected to. Plus with a flat pedal, you have your shoes possibly moving on the pedal. Plus there is the possibility of confirmation bias: you (actually anyone) feel that one bike is pedaling a bit strange, so you might be more inclined to interpret an indeterminate sensation as consistent with your hypothesis.

In your center diagram, the issue would be that the pedal hole is not drilled straight in the crank. I would assume that to be extremely rare on even the cheaper Shimano components. We have drilled holes into crankarms for many years now, so the industrial processes there are pretty settled.

In your third diagram, this should be extremely difficult to happen in real life. It would depend on what type of crank you have. If you have a crank where the spindle is bonded to one crankarm, then this would mean that spindle wasn't bonded straight to one arm. Again, this would be a manufacturing defect that should be very rare. Or if you have a traditional BB with a spindle and the cranks bolt to the spindle, then the mounting hole in one crankarm might have been broached crooked.

Not illustrated, but it's possible that the bottom bracket wasn't welded straight into the frame, so your cranks are pedaling off the frame's centerline. Again, this shouldn't happen. It could, maybe on a department store frame.

Bike stores should have the basic tools to check frame alignment. I don't believe I can clearly diagnose the issue here. If someone else doesn't do so, you could go into a bike store. In your place, I would probably tell them that I feel like the pedal isn't perpendicular to the crankarm and ask them to check.


When I felt a similar sensation on a tandem I'd not long bought second hand, I dismissed it as my shoe moving on the studded pedal, because I've had that before. 5km later the pedal fell off. The previous owner had put the right pedal in the left crank and vice versa.

In between I did actually grab the pedal and check for looseness, but hand force wasn't enough and it was dark and we were in a hurry.

So be careful swapping components for testing, and do check the threads are in good condition.

Cross threading is another similar possibility, as the pedal wouldn't be screwed straight into the crank arm (probably ruining the crank). This is be visible on close inspection - look from y the pedal side and the back of the crank (checking for L or R markings while you're at it).

You say nothing is loose, but does the pedal feel too tight or rough? Pedal bearings that are failing or in need of a service can jump under load.

I'm always wary of trying to measure angles on bike components. Many are cast or moulded, and have draw angles that mean you don't have a square face to measure against. In theory you can often cancel this out, but in practice it introduces an error comparable to the effect trying to measure.

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.