The Training Max is the number every 5/3/1 percentage is calculated from, and it is deliberately not your real max: Jim Wendler sets it at 90% of your true or estimated 1RM. This is the single most misunderstood part of the program. Lifters plug their gym-PR into the spreadsheet, wonder why week 3 buries them, and conclude 5/3/1 is too hard - when the program was never meant to run off the 1RM at all.
The 10% buffer is the engine of the program. Because the top sets are calculated from a submaximal number, you can add reps to the AMRAP sets week after week and add weight to the TM cycle after cycle without grinding. Wendler's reasoning: strength is built over years, and a program you can progress on for 12 months beats one that feels heavy for 6 weeks and stalls. Bar speed stays crisp, technique holds under fatigue, and the PRs come from beating rep records, not from testing maxes.
You do not need to test a true max to start 5/3/1. Two options:
The estimate is usually the better choice - it reflects what you can actually do for reps, which is what 5/3/1 trains. Hardy's free 1RM calculator does this math for you, and the training max guide walks through the whole calculation step by step.
| Week | Set 1 | Set 2 | Set 3 |
|---|---|---|---|
| Week 1 | 5 x 58.5 kg (65%) | 5 x 67.5 kg (75%) | 5+ x 76.5 kg (85%) |
| Week 2 | 3 x 63 kg (70%) | 3 x 72 kg (80%) | 3+ x 81 kg (90%) |
| Week 3 | 5 x 67.5 kg (75%) | 3 x 76.5 kg (85%) | 1+ x 85.5 kg (95%) |
| Week 4 (deload) | 5 x 36 kg (40%) | 5 x 45 kg (50%) | 5 x 54 kg (60%) |
Notice the week 3 top set is 85.5 kg - a weight a 100 kg bencher can press for 3-5 reps. That is the design working: the "1+" set should produce multiple reps, not a max-effort single.
Wendler's advice for your first cycle is blunt: start too light. If you are torn between two Training Maxes, take the lower one; some lifters even start at 85% of 1RM instead of 90%. A light first cycle costs you nothing - the AMRAP sets let you express whatever strength you have anyway - while a heavy first cycle costs you the next three months of clean progression. You will pass your old numbers within two or three cycles either way.
After each 4-week cycle, add 2.5 kg (5 lbs) to your press and bench TMs and 5 kg (10 lbs) to your squat and deadlift TMs. That is the whole progression model - small, almost insultingly patient increments that compound into big totals over a year of progressive overload. When a lift stalls - repeated missed reps or grinding AMRAP sets - reset that lift's TM to 90% of its current value (or recalculate it from a recent AMRAP estimated 1RM) and rebuild. Stalling and resets are covered in depth in 5/3/1 common mistakes.
Take your true or estimated 1RM and multiply by 0.9. A 140 kg squat 1RM gives a 126 kg Training Max, and every work set in the cycle is a percentage of that 126 kg.
Use weight x reps x 0.0333 + weight. If you press 60 kg for 9 reps, that is 60 x 9 x 0.0333 + 60 = about 78 kg. This is the formula Wendler uses throughout 5/3/1, and it is most accurate under about 10 reps.
Use a recent rep set to estimate each 1RM, take 90% of that for the TM, and if in doubt round down. Wendler would rather you start 5 kg too light than 2.5 kg too heavy - the program is built to progress for months, not to feel maximal in week one.
No. Every prescribed percentage in 5/3/1 - work sets, warm-ups, Boring But Big supplemental sets - is a percentage of the Training Max. The 1RM only exists to calculate the TM in the first place.
Hardy's free 5/3/1 Boring But Big routine derives your Training Max from a 1RM you enter or estimates it from a logged AMRAP set - then calculates every work set, rounded to your plates, for every cycle.
Get the RoutineDownload Hardy