You find a program that says “4 × 8 at 75%”. To run it, you need a 1RM. So you either go test it, or you plug numbers into a formula you found online. Either way you end up with a figure — and that figure doesn’t answer the question you’re actually asking: what do I load on the bar this morning, for this set?
This article covers what the 1RM really measures, why it’s a poor starting point for most lifters, and what goes in its place.
What a 1RM actually measures
Your 1RM (one-repetition maximum) is the heaviest load you can lift once, on a given exercise, with correct execution. It’s a measure of maximal strength. Nothing more, nothing less.
It’s performance data. The trouble starts when it gets used as prescription data — when “I can lift 100 kg once” becomes “therefore I train at 75 kg for 8 reps”. Those two statements don’t carry the same certainty.
The test is reliable. What follows isn’t.
Let’s kill a bad argument first, because it’s everywhere: “the 1RM test isn’t reliable”. That’s false.
Grgic and colleagues published a systematic review in Sports Medicine – Open (2020) on the test–retest reliability of the 1RM. For multi-joint exercises, the median intraclass correlation coefficient is 0.98, and the median coefficient of variation 4.3%. In untrained lifters, median ICC of 0.97. Translation: test your 1RM today, retest next week under the same conditions, and you land on roughly the same number.
The test measures what it claims to measure. The error isn’t in the measurement. It’s in what gets inferred from it.
70% of your 1RM is not 70% of mine
Here’s the real problem, and it’s brutal.
Richens and Cleather (2014, Biology of Sport) had strength-trained athletes and endurance runners perform a leg press set to failure at 70% of their respective 1RM. Same relative intensity. Same exercise.
| Profile (leg press, 70% of 1RM) | Reps to failure |
|---|---|
| Strength-trained athletes | 17.9 ± 2.8 |
| Endurance runners | 39.9 ± 17.6 |
More than double. And look at the standard deviation in the second group: ± 17.6. Some subjects did more than twice as many reps as other subjects within the same group, at the same relative intensity.
Add the exercise effect on top: at equal relative intensity, you’ll grind out far more reps on a leg press than on a bench press. Muscle mass involved, stability demands, systemic fatigue — a percentage captures none of it.
The conclusion is unavoidable: “4 × 8 at 75%” can be a comfortable RIR 4 set for you and a set taken to failure for the person next to you. The percentage doesn’t tell you the effort. And effort is what drives adaptation — not the percentage.
The estimation formulas: how far to trust them
If you don’t test, you estimate. The two most common formulas:
- Epley — 1RM = load × (1 + reps ÷ 30)
- Brzycki — 1RM = load ÷ (1.0278 − 0.0278 × reps)
Take someone pressing 80 kg. Here’s what each formula predicts, depending on how many reps were performed:
| Set performed | Epley | Brzycki | Gap |
|---|---|---|---|
| 80 kg × 5 | 93.3 kg | 90.0 kg | 3.3 kg |
| 80 kg × 10 | 106.7 kg | 106.7 kg | 0 kg |
| 80 kg × 15 | 120.0 kg | 131.0 kg | 11 kg |
The table says it all. The formulas agree around 10 reps — which is exactly where you don’t need them, since that’s already the range you train in. They diverge by 11 kg as soon as you move away from it. And both assume a universal load-to-reps relationship. Richens and Cleather just showed that relationship doesn’t exist.
A reasonable estimate up to 5 or 6 reps. A guess beyond 10.
And your Tuesday 1RM isn’t your Saturday 1RM
The test is repeatable under the same conditions. Except your conditions are never the same. A short night, a rough week at work, a leg session three days ago, a skipped meal — your maximal strength on the day moves.
So when you apply 75% to a number measured six weeks ago, you’re stacking two approximations: a percentage that doesn’t predict your effort, applied to a value that has since shifted. The result isn’t wildly wrong. It’s just permanently fuzzy. And that fuzziness is what makes you under-load or over-load, session after session, without noticing.
So — should you test it?
Three honest answers, depending on who you are.
Yes, if the 1RM is the goal. Powerlifting, weightlifting, strength sport: the one-rep maximum is the performance. Measuring it isn’t a detour — it’s the target. You should test it, and regularly.
Yes, occasionally, if you’re advanced and supervised. On a lift you’ve owned for years, with a spotter, two or three times a year, as a long-term progress marker. That’s legitimate.
No, in just about every other case. You burn an entire session, you accumulate neural fatigue that will drag on the next two, you take a non-zero risk under a heavy bar — for a number that expires, that doesn’t predict your rep count, and that still doesn’t tell you what to load this morning.
If you train for muscle, the case collapses even faster. Schoenfeld et al. (2017) is clear: as long as sets are taken close to failure, heavy and light loads produce comparable hypertrophy. It’s maximal strength that demands heavy loads. Chasing a precise percentage to build muscle is optimising the wrong variable.
What to use instead: calibrate, don’t test
The useful question was never “what’s my maximum”. It has always been “what load puts me in the right effort zone, today, on this exercise”. You can answer that in three steps, without ever going near a maximum.
1. Find the right neighbourhood
Ramp up through your warm-up. Stop when you feel that at this load, over your target rep range, you’d have 3 or 4 reps left in the tank. You don’t need to be exact. You need to be in the right neighbourhood — precision comes on its own.
2. Report the actual effort
At the end of each working set, log your RIR: how many reps you had left before technical failure. That’s a data point from today, not an extrapolation.
3. Let the rule decide
- RIR ≥ 3 across all working sets → the load is too light. Go up.
- RIR 1 to 2 → you’re in the zone. Hold, and add reps.
- RIR 0, or reps you can no longer complete → hold, or back off a notch.
Two or three sessions are enough for the load to settle. No max test, no formula, no percentage. And crucially: the adjustment happens every single session. A 1RM is a photograph. This is a video.
That’s exactly the logic RyzeFit is built on — the reason the app never asks you how much you can lift.
If you insist on testing: the clean protocol
Still want to do it? Fine. Then do it properly.
- Only on a lift you’ve owned for months. Never on a movement you learned three weeks ago.
- Full warm-up, then ramp: roughly 5 reps at 50%, 3 reps at 70%, 1 rep at 85%, then your attempts.
- 3 to 5 attempts maximum, with 3 to 5 minutes of rest between each. Beyond that, fatigue corrupts the measurement — you stop testing strength and start testing your endurance at maximal effort.
- A spotter is mandatory on bench and squat, or safety pins set at the right height.
- This isn’t a session. It IS the session. Nothing else afterwards.
⚠️ Never test a 1RM at the end of a session, or during a week where you’re already fatigued. You’d get a low number, anchor your percentages to it, and train too light for six weeks.
Honest limitations
Three things worth putting on the table.
RIR isn’t free either. It takes learning. Beginners almost always overestimate their margin — they call RIR 2 when they were at 0. It’s bodily knowledge, built over months.
The 1RM has real value. It’s a shared language between lifters, a progress marker across years, and for many a legitimate source of motivation. This article argues it’s a poor day-to-day prescription tool — not that it’s useless.
There is a third path. Velocity-based training measures effort objectively, without depending on your perception. It’s more rigorous than RIR. It’s also hardware, budget, and a constraint almost nobody has in a commercial gym.
Further reading
- RIR (Reps In Reserve) Explained — the metric that replaces 1RM percentages, and how to estimate it without lying to yourself.
- How to Increase Weight in the Gym — the concrete progression rules once the load is calibrated.
- Progressive Overload: The Complete Guide — the mechanism that actually matters, rather than the maximum figure.
- Why You’re Not Gaining Muscle — including mis-set intensity, which is exactly what blind percentages produce.
- How Many Sets Per Week for Muscle Growth — because load is only one variable; volume is another.