Training Max vs One Rep Max: The 90% Rule
Your one rep max is what you can lift on your best day. Your training max is what you build your programme on — typically 90% of the real thing. That deliberate 10% discount, popularised by Jim Wendler's 5/3/1, is one of the most useful ideas in strength training. Here is why it exists, how to set it, and the mistakes people make with it.
The two numbers, defined
- One rep max (1RM): the heaviest weight you can lift once with good form — tested, or estimated with the 1RM calculator from a hard set.
- Training max (TM): a deliberately conservative working number, usually 85–90% of your 1RM, from which all your programme percentages are calculated.
When a 5/3/1 sheet says "85%", it means 85% of the training max — which is about 76% of your true max. That single layer of discounting is the whole trick, and the homepage calculator shows your 90% training max with every estimate for exactly this reason.
Why train from 90% instead of 100%?
Bad days exist. Your tested max was earned fresh, motivated, and probably on a Saturday. Your Tuesday session after work is a different athlete. Percentages of a true max assume best-day strength every day; percentages of a training max build in the buffer real life requires — the top set stays achievable, and an achievable top set gets done with good form.
Bar speed stays honest. Submaximal weights move fast, and fast reps with perfect positions build strength and skill together. Grinding every top set at the edge of your ability teaches grinding.
Progress becomes boring — in a good way. The training max rises in small fixed steps (typically +2.5 kg for upper-body lifts and +5 kg for lower-body lifts per 3–4 week cycle in 5/3/1). Small steps, taken relentlessly, compound. Lifters who anchor to their true max tend to stall, miss reps, and reset; lifters who anchor to a training max tend to still be progressing a year later.
A worked example: squat, 1RM 200 kg
| From true 1RM (200) | From training max (180) | |
|---|---|---|
| "65%" day | 130 kg | 117.5 kg |
| "75%" day | 150 kg | 135 kg |
| "85%" day (top set) | 170 kg | 152.5 kg |
Look at that top set. 152.5 kg is a weight the lifter can hit for a strong set of 5 — hard, productive, repeatable. 170 kg for reps, demanded on an ordinary weeknight, is where form breaks and cycles die. The training max column looks "too light" on paper and outperforms the true-max column in the gym, month after month. Our article on training percentages shows how those percentages map to goals once the base number is sane.
How to set and adjust your training max
- Setting it: take a recent, honest 1RM — tested or estimated from a set of 3–5 — and multiply by 0.9. If your estimate came from a sketchy high-rep set, use 0.85 or re-estimate first; the TM inherits every error in the 1RM.
- Raising it: only at the end of a completed cycle, and only by the small fixed steps. Never raise it mid-cycle because a set felt easy — that feeling is the system working.
- Lowering it: if you miss reps on the top set in two consecutive cycles, your training max is fiction. Drop it 10% without drama and rebuild. A training max you cannot lift is not a training max; it is a souvenir.
- Cross-checking with effort: if your top sets consistently rate far below target effort, your TM may be drifting low relative to your strength — the fix is the scheduled end-of-cycle bump, not a mid-cycle leap. Combining percentages with honest effort ratings, as in our guide to using RPE in training, catches drift in both directions.
The classic mistakes
Using your lifetime best as the base — a max from two years and 6 kg of bodyweight ago is history, not data. Setting the TM at 100% "because the percentages already look light" — they look light because the discount is the point. And testing a new max every cycle to reset the TM: that converts a building programme back into a measuring programme. Recalculate or test at sensible intervals, reset the TM from the new number, and get back to the quiet work of adding 2.5 kg at a time.
Frequently asked questions
Is a training max always 90% of 1RM?
90% is the standard starting point (5/3/1 uses 90%, some systems use 85%). Lifters who know their max is an optimistic estimate — from a high-rep set, or a very good day — should start at 85%. The exact figure matters less than the principle: the base must be a weight you could own on an average day.
Won't training from 90% make me weaker?
No — you still train hard; the top sets are challenging sets of 3–5, not warm-ups. What changes is that the loads are repeatable, so you accumulate far more quality work across months. Lifters routinely set genuine new maxes after cycles built entirely on a 90% training max.
How often should I reset my training max from a new 1RM?
Every few cycles, or when the fixed bumps no longer reflect reality — typically every 3–6 months. Test or recalculate, take 90% of the fresh number, and continue. Frequent resetting defeats the purpose; the steady small increases are the engine.