Guide to One Rep Max
What is One Rep Max?
One Rep Max (1RM) is the maximum weight you can lift in a single repetition while maintaining proper form. It is a fundamental parameter in strength training used for planning workout intensity and programming loads. Knowing your 1RM allows you to determine what weights to use at different repetition ranges, enabling systematic progress in training.
1RM percentages in training
- 100% - maximum weight, 1 repetition, used for testing
- 85-90% - maximal strength, 2-4 repetitions
- 75-80% - strength + mass, 6-8 repetitions
- 65-70% - hypertrophy, 8-12 repetitions
- 50-60% - endurance, 15+ repetitions
Calculation formulas
The most popular formulas are Epley (weight × (1 + reps/30)) and Brzycki (weight × (36 / (37 - reps)). The Epley formula is most commonly used and gives good results in the 1-10 rep range. Remember results are estimates - actual 1RM may vary depending on fitness, form, and exercise type.
Safety and testing
Testing 1RM should be done carefully, preferably with a spotter. Warm-up is essential - perform sets with decreasing weights before the max attempt. Don't test 1RM more often than every 4-6 weeks as it requires significant nervous system effort. For most training goals, calculating 1RM based on weight to failure is safer and sufficient.
One-rep max formulas agree at five reps and fall apart at twenty
Estimating a one-rep max from a heavier set of several reps is a curve fit, not a measurement. The common formulas were fitted to low-rep data and behave well there. Push them past about ten reps and they diverge wildly — at twenty reps the three standard formulas disagree by more than 75 kg on the same lift.
How it works
- Estimates a one-rep max from a weight and the number of reps completed with it.
- Applies the Epley, Brzycki and Lombardi formulas so the spread between them is visible.
- Makes the reliable range explicit, which matters more than any single number the tool produces.
Epley 1RM = w × (1 + reps ÷ 30) Brzycki 1RM = w × 36 ÷ (37 − reps) Lombardi 1RM = w × reps^0.10 all three are curve fits to observed data, not physiology Brzycki divides by zero at 37 reps
Worked example
Lifting 100 kg, and estimating the one-rep max from different rep counts.
- 3 reps → Epley 110.0, Brzycki 105.9, Lombardi 111.6 — spread 5.7 kg
- 5 reps → 116.7, 112.5, 117.5 — spread 5.0 kg
- 10 reps → 133.3, 133.3, 125.9 — spread 7.4 kg
- 15 reps → 150.0, 163.6, 131.1 — spread 32.5 kg
- 20 reps → 166.7, 211.8, 134.9 — spread 76.8 kg
Below about eight reps the three formulas land within a few kilos of each other. By twenty reps they disagree by 76.8 kg, which is more than three quarters of the weight actually lifted. The estimate stops meaning anything long before the arithmetic stops working.
Reading the result
- Use sets of two to six reps if you want a usable estimate. That is the range the formulas were fitted to and where they broadly agree, which is the only real evidence that any of them is right.
- Epley does not return the lifted weight at a single rep — it reports 103.3 kg for a true 100 kg single. Brzycki and Lombardi both collapse correctly to the weight itself. Worth knowing if you are comparing an estimate against a genuine max.
- High-rep sets measure muscular endurance more than maximal strength, and the two are only loosely related. That is the underlying reason the formulas diverge rather than a flaw in any one of them.
- An estimate is not a target. Attempting a true single at a calculated number, without building up to it, is how people get injured — particularly on lifts where failure is difficult to escape.
Common questions
- Which formula should I use?
- In the two-to-six rep range it barely matters — they agree within a few kilos. If you want one default, Epley is the most widely used; just remember it overstates slightly at a single rep.
- Why do they disagree so much at high reps?
- Because they were fitted to low-rep data and extrapolate differently beyond it. Brzycki's denominator shrinks toward zero as reps approach 37, so it climbs steeply; Lombardi's power curve flattens. Neither behaviour is grounded in evidence out there.