How accurate are 1RM calculators?

1RM calculators are generally accurate within 5-10% when used with rep ranges under 10. The accuracy depends on factors like training experience, muscle fiber composition, and the specific formula used. Our calculator uses 7 different formulas to give you the most comprehensive estimate.

Which 1RM formula is most accurate?

The Epley formula is most widely used and accurate for most people, especially with rep ranges of 2-10. However, different formulas work better for different individuals. Brzycki tends to be conservative, while Lombardi works well for powerlifters. Our calculator shows all 7 formulas so you can compare.

Can I use this for any exercise?

Yes, you can use 1RM calculators for any compound exercise like bench press, squat, deadlift, overhead press, and rows. The formulas are based on general strength principles that apply across exercises, though they're most accurate for compound movements.

What rep range should I use for the most accurate calculation?

For best accuracy, use rep ranges between 2-10 reps. The sweet spot is 3-8 reps where most formulas perform optimally. Avoid using very high rep ranges (15+) as the formulas become less reliable due to muscular endurance factors.

How often should I recalculate my 1RM?

Recalculate your 1RM every 4-6 weeks or whenever you notice significant strength gains. For beginners, monthly calculations work well. Advanced lifters might recalculate every 6-8 weeks. Always use recent max effort sets for the most accurate results.

Why do 1RM formulas give different results?

Each formula was developed using different study populations, methodologies, and underlying mathematical structures (linear, hyperbolic, exponential, logarithmic). They make different assumptions about how strength endurance falls off with reps, which produces different predictions especially at the extremes (very low and very high rep counts). Read more in our deep dive: Why Different Formulas Give Different Results.

Should I pick one formula or use an average?

Use a trimmed-mean average across multiple formulas for the most reliable estimate. Single formulas have known biases — Lander underestimates, Lombardi underestimates for beginners, Epley overestimates at high reps. Averaging cancels out biases. Trimming (removing the highest and lowest estimates) further protects against any one outlier formula skewing the result.

Which formula is most conservative?

Lander produces the lowest estimates of the seven formulas in most rep ranges, followed by O'Conner. Both are intentionally conservative — designed for safety-focused programming where overshooting your true 1RM creates injury risk. If you want the lowest reasonable estimate, average Lander and O'Conner.

Which formula gives the highest 1RM estimate?

At low reps (1-3), Epley and Brzycki tend to produce the highest estimates. At high reps (10+), Epley overshoots significantly while exponential models (Mayhew, Wathan) stay accurate. If a formula is producing visibly higher numbers than the others, it usually means the formula's assumptions don't fit your rep range — switch to a more rep-range-appropriate model.

Formula Hub · 7 formulas compared

1RM Formula Comparison

All seven major 1RM prediction formulas — equations, accuracy data, study citations, and head-to-head comparisons. The page that answers “which 1RM formula is the most accurate” with research-grounded specifics.

All 7 formulas at a glance

Formula	Year	Equation	Best for	Accuracy	Bias
Epley	1985	1RM = weight × (1 + reps / 30)	2-10 reps	±5%	balanced
Brzycki	1993	1RM = weight × 36 / (37 − reps)	1-10 reps	±3% for reps under 10	balanced
Lombardi	1989	1RM = weight × reps^0.10	3-12 reps	±4% for trained individuals	underestimates
O'Conner	1989	1RM = weight × (1 + reps × 0.025)	1-12 reps	±6% across all rep ranges	underestimates
Mayhew	1992	1RM = 100 × weight / (52.2 + 41.9 × e^(−0.055 × reps))	1-15 reps	±4% consistently	balanced
Wathan	1994	1RM = 100 × weight / (48.8 + 53.8 × e^(−0.075 × reps))	1-12 reps	±3% for heavy loads	balanced
Lander	1985	1RM = 100 × weight / (101.3 − 2.67123 × reps)	1-10 reps	±5% with built-in safety margin	underestimates

Click any formula name to see its history, calculator, validation data, and FAQ.

Side-by-side: same set, all 7 formulas

How much do the formulas actually disagree? Here are four sample sets run through every formula. Notice the spread tightens at low reps and widens at higher reps.

Set	Epley	Brzycki	Lombardi	O'Conner	Mayhew	Wathan	Lander
135 lb × 10 spread: 13 lb	180	180	170	169	177	182	181
225 lb × 5 spread: 15 lb	263	253	264	253	268	262	256
315 lb × 3 spread: 25 lb	347	334	352	339	359	343	338
405 lb × 1 spread: 36 lb	419	405	405	415	441	410	411

Amber = highest estimate · Blue = lowest estimate. The trimmed-mean average (used by our main calculator) drops the highest and lowest, then averages the middle five for stability.

Pick the right formula for your situation

Testing 1-3 reps (heavy singles/doubles)

Use Wathan or Brzycki. Both are calibrated on heavy training data and produce the most accurate predictions when reps are limited.

Testing 4-8 reps (most common)

Almost any formula works — they converge in this range. Use Epley for fast back-of-envelope math or the trimmed-mean average from our main calculator.

Testing 10+ reps (high-rep sets)

Use Mayhew — its exponential structure stays accurate at high rep counts where linear formulas (Epley, O'Conner) start to overestimate significantly.

Beginner or returning lifter

Use Lander or O'Conner. Both intentionally underestimate, which produces safer programming percentages.

Experienced lifter testing accurately

Use Lombardi or Wathan. Both are calibrated on populations with high neuromuscular efficiency.

Validation research

LeSuer et al. (1997)

Mayhew, Wathan, and Brzycki produced the smallest error across bench, squat, and deadlift. Most formulas underestimated 1RM as reps increased.

LeSuer, D.A., McCormick, J.H., Mayhew, J.L., Wasserstein, R.L., & Arnold, M.D. (1997). The accuracy of prediction equations for estimating 1-RM performance in the bench press, squat, and deadlift. Journal of Strength and Conditioning Research, 11(4), 211-213.

Reynolds et al. (2006)

Standard error of estimate ranged from 5.6 to 7.9 kg across formulas; high reps (>10) reduced accuracy meaningfully.

Reynolds, J.M., Gordon, T.J., & Robergs, R.A. (2006). Prediction of one repetition maximum strength from multiple repetition maximum testing and anthropometry. Journal of Strength and Conditioning Research, 20(3), 584-592.

Wood et al. (2002)

In older adults, most formulas overestimated 1RM by 5-15%; conservative choices (Lander, O'Conner) outperformed.

Wood, T.M., Maddalozzo, G.F., & Harter, R.A. (2002). Accuracy of seven equations for predicting 1-RM performance of apparently healthy, sedentary older adults. Measurement in Physical Education and Exercise Science, 6(2), 67-94.