How accurate is a one-rep max calculator?

A calculator that turns your set of eight into a one-rep max is doing something quietly heroic: collapsing every exercise, every body, and every load into one curve. The biggest synthesis we have says that curve does not exist. A meta-regression in Sports Medicine pooled 952 repetitions-to-failure tests from 7,289 people across 269 studies and found that at 80 percent of 1RM, the average lifter managed about 8.8 reps on the bench press but about 13.1 on the leg press (see [1]). Same percentage, four extra reps. Push both through the same formula and one of those estimates is going to be badly wrong.

The newer wrinkle is that the load itself bends the curve. A 2026 preprint analyzing 303,494 near-failure sets across 388 exercises logged in a training app found the conversion factor between reps and maximal capacity depends logarithmically on the weight on the bar, so that at light weights each additional rep implies a larger fraction of your maximum than it does at heavy weights (see [2]). Building that weight dependence into the equation cut prediction inconsistency by 17 to 22 percent against four classical formulas, and 91 percent of the improvement came from the weight term alone. Epley and Brzycki were never wrong so much as too simple.

Here is the part that should make you feel better about your own numbers. Sex, age, and training status did not clearly move the reps-versus-percentage relationship in that meta-regression (see [1]). The error is not personal. It lives in which lift you picked and how heavy the bar was. A 2024 study following 119 young adults through 16 weeks of training came at it from the other side and found the best-performing equation changed by lift, with bench press and squat predicting cleanly while arm curl estimates were the least reliable of the three (see [3]).

So use the calculator, just stop treating it as a measurement. Estimates taken from sets of five or fewer on a lift you actually train sit closest to the truth, and they are most useful compared against your own previous estimate on that same exercise rather than across lifts, which is the case our guide to estimated 1RM makes in more detail. If your bench estimate climbs ten pounds across a block while your rep quality holds, that is real progress whether or not the number would survive a true single. Tracking direction without burning a week recovering from a max attempt is the entire job of the metric.