How Accurate Are 1RM Calculators?

Short answer: accurate enough to program your training, provided you feed them the right set. A hard set of 3–5 reps gives an estimate typically within 2–5% of your true max. A set of 15 gives a number you should treat as a rough guess. Here is why accuracy depends so heavily on the rep range — and how to get the most trustworthy estimate.

Accuracy by rep range

Set you enter	Typical accuracy	Verdict
2–3 reps	Within ~2%	Excellent — barely an estimate at all
4–6 reps	Within ~2–5%	The sweet spot for most lifters
7–10 reps	Within ~5–8%	Usable, but individual variation grows
11–15 reps	Within ~10% or worse	Rough guide only
15+ reps	Unreliable	Do not program from this number

Take a concrete example. You bench 100 kg for 5 reps. The Epley formula gives 100 × (1 + 5 ÷ 30) = 116.7 kg, and our one rep max calculator averages six formulas to land in the same neighbourhood. A lifter who can genuinely bench 100 × 5 will, on a good day, usually single somewhere between 112 and 120 kg. That band is plenty accurate for choosing training weights — the difference between programming off 114 or 118 kg is a rounding error on the bar.

Why high-rep estimates go wrong

Every 1RM formula assumes a fixed relationship between reps and percentage of max — roughly, that each extra rep costs you about 2–3% of the load. That relationship holds well at low reps. At high reps it breaks down, because the limiting factor changes. A set of 15 is no longer limited mainly by maximal strength; it is limited by muscular endurance, metabolite tolerance, pacing and plain willingness to suffer. Those qualities vary far more between people than strength does.

Two lifters can both squat 140 kg. One is a former rower with huge endurance and gets 15 reps at 100 kg; the other is a pure strength athlete and gets 9. The formula gives them wildly different "maxes" from those sets — and both estimates are fiction. The formulas also start disagreeing with each other at high reps, which is exactly what our Epley vs Brzycki comparison shows with worked numbers: at 5 reps the formulas nearly agree; at 12 they are several kilos apart.

The other things that bend the number

• The lift itself. Estimates are most reliable for the squat and bench press. Deadlift estimates skew because grip, starting position and the lack of an eccentric phase change the rep relationship. Upper-body isolation lifts are the least reliable of all.
• Training status. Beginners often underperform their estimate on a true single — the skill of grinding a maximal attempt is not there yet. Experienced lifters usually land close to the prediction.
• How the set was done. Bounced reps, shortened range of motion and long rests between reps ("rest-pause" in disguise) all inflate the input and therefore the estimate. Honest reps only.
• Fatigue. A set taken after heavy work earlier in the session underestimates you. Take your test set fresh, early in the workout.

How to get the most accurate estimate

Warm up properly, then take one hard set of 3–5 reps to within a rep of failure with clean technique, early in a session, on a lift you practise regularly. Run it through several formulas and use the average. Then sanity-check it: the estimate should feel believable relative to your training weights. If you regularly handle 85% of the predicted max for triples, the number is in the right place. And when the number really matters — a competition or a milestone attempt — there is no substitute for testing properly, as covered in our guide to testing your one rep max safely.

Frequently asked questions