I've been fixing up my wife's long-neglected and heavily-ridden commuter and have realized that the rear wheel sits way closers to the non-driveside seat stays than to the driveside. The wheel sits perfectly square between the chain stays. Frame is steel with horizontal dropouts.

Any idea where I should start checking for frame alignment issues? Or does anyone have any other suggestions?

If your wife isn't reporting any riding or handling issues, the place to start is with a stiff shot of your favorite drink, and forget about it.

There are countless possibilities, running from an incorrectly dished rear wheel (if you have horizontal dropouts) to the frame having been built this way all along.

This is the kind of thing that rarely changes, so odds favor it being this way all along unless the bike was in a collision.

If you choose to obsess over this, start with removing and remounting the rear wheel flipped, (cassette on left) and seeing what, if anything changes (test for dish). Also, keeping the wheel in the frame with the bike vertical, try slowly rotating the axle and seeing it the top of the wheel moves to either side as you so so (test for bent axle).

If the wheel passes both tests, it's a frame issue and you're best off leaving it be UNLESS it is affecting the ride.

OK, since someone has to ask: how much closer (in mm)?

horizontal dropouts offer axle slipping possibilities..

done the string test?







....

A pretty significant amount -- around 10mm. There have never been any complaints about function but it does make setting up the brakes fairly strange.

Horizontal dropouts but no amount of adjustment compensates for this.

I've had this on two bikes. I suspected that it was NOT wheel dish because, like yours, the wheel was centered in the chain stays. I checked dish and it was good. I checked frame alignment using the string method, and it was good.

With the frames stripped down to bare naked I discovered what was wrong. Setting the frame on a known flat surface, it would teeter-totter. Meaning, if the bottom bracket was flat on the deck, then one dropout would be touching the deck and one was a few MMs off the deck. If I pushed both dropouts down to make contact with the deck, the bottom bracket didn't sit level. Upon further and closer inspection, and measuring, I could see that one dropout was positioned and welded into the stays slightly higher than the other. I confirmed this by measuring chain stay heights with the bottom bracket flush and level. It was just the dropouts.

These were cheap bikes so my solution was very easy. I took a file to the "low" dropout. Removing a tiny bit at the dropout has a much bigger affect at the top of the wheel. Got it sorted very quickly, easily, and it's not visible because the QR or axle nut will hide the work.

Horizontal dropouts, in combination with an incorrectly dished wheel makes it easy to center in the chainstays while the error will remain at the seatstays.

Go back and perform the two wheel tests I described earlier and report back.

However, keep in mind that since you can always center in the chainstays with horizontal dropouts, it's very possible that the triangle is offset to either side. Do the dish test and report back, then we can look at various frame checks.

In case someone ever finds this thread, the problem was that I put the spacers back on the wrong side of the wheel after servicing the hub. Took the wheel apart, flipped things around and now it works great. This is the kind of dumb thing that you do when you get too used to single speeds, I guess.

Well done.