health · 2026-05-01
Estimate 1RM from a sub-max set using Brzycki, Epley, and Lombardi formulas — averaged for a robust prediction.
| Weight lifted (lb) | 225 |
| Reps completed (RIR 0) | 5 |
| Brzycki 1RM | 253 |
| Epley 1RM | 263 |
| Lombardi 1RM | 264 |
| 80% training weight | 208 |
| 70% training weight | 182 |
Three classic 1RM estimators, each calibrated against different rep ranges:
weight × 36 / (37 - reps) — accurate at 2-10 reps, breaks down outsideweight × (1 + reps/30) — slightly conservative, smoother across rep rangesweight × reps^0.10 — power curve, holds up better at higher repsThe arithmetic mean of all three is more robust than any single formula. At 3-6 reps the spread is tight (within 2-3%); at 10+ reps the formulas diverge and accuracy drops.
These formulas assume standard barbell movements (squat, bench, deadlift, OHP). They overshoot on high-velocity lifts (cleans, snatches) and undershoot on slow grinders (deadlifts done to lockout from a dead stop).
Track 1RM trends, not absolute values. A 5 lb week-over-week increase in estimated 1RM means progress; the absolute number matters less than the slope.
RIR 0 (zero reps in reserve) is your true rep limit. For accurate 1RM estimates you need this — leaving 1-2 in the tank inflates the estimate by 5-10%. Use spotters, safety pins, or stop sets you're not confident about. Better an underestimate than a missed rep with bad form.
Each was calibrated against different lifter populations and rep ranges. Brzycki overshoots at 1-3 reps, Epley undershoots at 8+ reps, Lombardi handles the middle. Averaging cancels the directional biases.
Once or twice a year, for confidence and meet prep. Otherwise use estimated 1RM updated from your weekly working sets. Frequent maxing trashes recovery, ego-tests weights you wouldn't program, and adds injury risk for marginal data gain.
2-10 reps is the sweet spot. Below 2, the formulas just multiply your weight by ~1.03; not useful. Above 10, accuracy drops fast — at 15 reps you're really measuring muscular endurance, not 1RM-relevant strength.