How to find your one-rep max: real test vs. estimating from reps and RPE, a sensible warm-up ramp and how often you should actually max out.
Your 1RM — the one-rep maximum — is the heaviest load you move once with clean, competition-valid technique. It is the currency of powerlifting: your three best 1RMs in squat, bench and deadlift make up your total.
The most honest answer up front: you do not have to truly max out constantly to know your 1RM. From a set of a few clean reps plus your RPE, the 1RM can be estimated reliably — and that is far gentler on your nervous system than a real grinder every week.
Why lift heavy at all? Because max strength only comes with heavy load: high loads are clearly superior to light ones for pure strength, and the 1–5 rep range develops strength most effectively. So if you want to get strong, you must lift heavy regularly — but "heavy" does not mean "to the absolute limit every week".
A max attempt is 90% warm-up work. Climb from the empty bar in stages — roughly 40% × 5, 60% × 3, 75% × 2, 85% × 1, 92% × 1 — with falling reps. This is RAMP logic (Raise, Activate, Mobilise, Potentiate): it primes your nervous system without fatiguing you before the heavy attempt. Skip long static stretching beforehand — it briefly lowers your output.
After the ramp comes one single heavy attempt with full competition technique. If it goes cleanly, you can optionally add a small jump (2.5–5 kg). No grinding with collapsing technique — an ugly grinder is not a valid 1RM and only an injury risk.
To spare your nervous system, do not test for real but extrapolate. Run a set of 3–5 clean reps and use the Epley formula: 1RM = load × (1 + reps / 30). A set of 5 with 120 kg gives about 140 kg estimated max. The closer the set is to failure, the more accurate the estimate.
A pure formula does not know how hard the set really was. So couple it with RPE: a set of 5 @ RPE 8 (2 RIR) corresponds roughly to a 7-rep max, not a true 5RM. RPE-based estimates hit the real 1RM surprisingly well. That way you get your number without ever going to the absolute limit.
A 1RM attempt is the moment of highest failed-rep risk. Safety pins and a spotter are mandatory, especially on bench. Never go for a real maximum if sleep, nutrition or technique are off that day — sleep loss demonstrably lowers your max strength and raises injury risk. And the simplest rule: if technique collapses mid-attempt, the attempt is over. An estimated 1RM from a clean multi-rep set is always worth more than a real one from a dangerous grinder.
Keep a rolling "e1RM" curve (estimated 1RM) in your log: from every heavy working set, compute an estimated 1RM via RPE and log it. That way you see your strength progress every week, not just on rare test days — and you spot early whether a program is working or you are heading for a plateau. Your e1RM curve is more honest than your gut feeling.
Rarely — roughly at the end of a training block (every 8–16 weeks) or in a meet. Strength is less volume-dependent than hypertrophy, and frequent maxing costs more recovery than it adds in progress. Day to day, estimating via RPE and formula is entirely enough to track your progress.
Fairly accurate when the set was near failure and had few reps — a 3 or 5 estimates better than a 12. The more reps, the less reliable the extrapolation. Couple the formula to your RPE and the estimate becomes much more realistic, because it accounts for how much was really left in the tank.
Because strength is specific: the nervous system learns exactly the task you give it. High loads develop max strength clearly better than light ones, and the 1–5 rep range is the sweet spot for it. Light loads to failure build muscle but not the same raw max strength.
Yes. Most of your strength gains come from heavy submaximal work (about 80–92%) over weeks, not from maxing itself. Many athletes lift nothing above RPE 9 for months and only hit the real limit in a meet — their estimated 1RM still rises steadily. Real maxing only confirms what training built.
Block or DUP: Building Your Strength Mesocycle Right