Match your symptom to the cause
| What happened | Most likely cause | Fix |
|---|---|---|
| Missed the lift at the exact predicted weight | Neural fatigue from warm-ups, or the estimation set was closer to failure than it felt | Rest 3-5 min between heavy warm-up sets; stop 1 rep short of failure on your test set |
| Actual max is well below the estimate | Estimation set had more than ~10 reps — formula error grows with rep count | Re-test with a 3-6 rep set for a tighter estimate |
| Actual max is well above the estimate | Formula bias (Brzycki underestimates high-rep sets) or high fast-twitch fiber ratio | Average multiple formulas, or weight Lombardi more for sets over 8 reps |
| Right on some lifts, off on others | Technique-limited lifts don't follow the same fatigue curve as simpler ones | Trust the estimate less for squat/bench, more for machine or single-joint lifts |
| Was accurate before, feels wrong now | The estimate is only as fresh as the set it came from — strength shifts with training phase | Recalculate from a recent set rather than reusing an old number |
Get a fresh estimate from your most recent working set.
1RM Calculator →Why the formulas diverge above 10 reps
Every 1RM formula (Epley, Brzycki, Lombardi, and others) is built on the same underlying assumption: that the relationship between reps performed and percentage of your true max follows a predictable curve. That assumption holds up well for sets of 3-8 reps. Past about 10 reps, muscular endurance and cardiovascular fatigue start contributing to when you fail the set, not just raw strength — and different people's endurance varies far more than their raw strength does. This is why a 15-rep set produces a much less reliable estimate than a 5-rep set, even though both are mathematically valid inputs.
Why some formulas run higher or lower than others
The most commonly used formulas — Epley and Brzycki — return identical results at exactly 10 reps by design, but diverge below that. Epley tends to estimate slightly higher for low rep counts; Brzycki tends to run more conservative, particularly for sets in the 6-9 rep range. Neither is universally "more correct" — the gap between them is often a reasonable proxy for your own uncertainty range. If your fast-twitch muscle fiber ratio is unusually high (common in explosive, power-oriented lifters), you'll tend to outperform most formulas' predictions, since these formulas are built from average population data.
Why technique-heavy lifts are the least predictable
Formulas were originally validated primarily on simpler, more mechanically consistent movements. Lifts with a technical component — squat depth requirements, a paused bench press, a deadlift off a deficit — introduce a variable the formula doesn't account for: how much of your "failure" on a submaximal set was strength-limited versus technique-limited. A leg press or machine chest press estimate will typically track much closer to your true max than a free-weight back squat estimate, simply because there's less technical variance to introduce error.
How to test a true 1RM safely, if you choose to
Only attempt it with a training history: Several months of consistent, technically sound training first — testing a true max as a beginner adds injury risk for little useful information.
Warm up in ascending steps: Several progressively heavier sets (roughly 50%, 70%, 85%, 95% of your estimated max) with full recovery between them, not a single warm-up set.
Use a spotter or safety bars: Non-negotiable for barbell squat and bench press attempts specifically.
Limit true max attempts to your top 1-2 lifts: Testing every exercise to a true max in one session compounds fatigue and lowers accuracy on the later lifts.
How often to retest
Every 8-12 weeks is a reasonable default for most structured training programs — roughly the length of a typical training block. Testing weekly adds cumulative fatigue without producing meaningfully different numbers, since strength doesn't change that quickly. Recalculate sooner than your usual schedule if you've just come off a deload week, returned from an illness or long break, or made a significant change in training frequency — any of these can shift your true number meaningfully outside the normal 8-12 week cycle.