Free strength standards tool · no signupYour lift.
On the platform.
Where a squat, bench, or deadlift would place at a drug-tested raw meet, by sex, bodyweight, and age. It compares you against people who compete. No honest dataset exists for the general population.
Meet results come in two categories, so those are the two the data supports.
Deadlift percentiles
| Percentile | Lift | × bodyweight | Meaning |
|---|
This is not a general-population comparison
Worth saying plainly, because it decides how you should read every number above.
Everyone in this data entered a sanctioned powerlifting meet. They paid an entry fee, made weight, and lifted under judging. That is a heavily self-selected crowd. A 200 kg deadlift at 95 kg bodyweight sits around the 30th percentile here, and would turn heads in most commercial gyms. Both things are true. Only the first one is what this page measures.
A low percentile here does not mean you are weak. It means you would place in the lower part of a field made up entirely of competitive powerlifters. If you have never competed, that is the expected result, and it answers a different question from "am I strong".
Why there is no comparison to regular gym-goers
Because no honest version of it exists. Nobody has taken a census of what people lift, and the datasets that get pressed into service fall into two camps:
- Competition records are accurate and judged, and they only cover competitors. Even the largest published normative study, Van den Hoek and colleagues' 2024 analysis of 809,986 entries, is competition data.
- Crowd-sourced app logs cover a much broader group, but they are self-reported, unjudged, unverifiable, and skewed toward people motivated enough to log every set.
Publishing a "general population percentile" would mean inventing the population. This page would rather tell you something narrower and true.
The two fields you can choose
Both are made of competitors, but they are different competitors.
- First meet takes each lifter's debut result. This is as close as the data gets to someone who trained in a gym and then entered something, since it strips out everything they gained from years of competing afterwards.
- Career best takes each lifter's best result ever, so you are up against experienced competitors at their peak.
The gap between the two is smaller than people expect, usually around ten percentile points, and that is informative in itself. The distance from a debut competitor to a seasoned one is much shorter than the distance from any competitor to someone who has never stepped on a platform.
What has been filtered out
- Raw results only. Every included row is marked Raw in OpenPowerlifting. Wraps, squat suits, and bench shirts are excluded.
- Drug-tested results only. Every included row is marked Tested in OpenPowerlifting. Untested results are excluded.
- Successful attempts only. Missed lifts appear as negative numbers in the source data and are dropped.
- Implausible rows. Entries outside generous bodyweight-ratio bounds are treated as transcription errors. That removed 115 rows out of four million, which tells you outliers were never what made this distribution look strong.
Why bodyweight and age are separate inputs
Strength scales with bodyweight, though not in a straight line. A 100 kg lifter is not twice as strong as a 50 kg lifter. Comparing within a weight band sidesteps the assumptions baked into coefficients like Wilks or DOTS.
Age matters more than most tables admit. Masters lifters make up a large share of meet entries, and pooling a 58-year-old with a 26-year-old produces a distribution that describes neither of them. Selecting your age band narrows the comparison to people in the same decade. Where a band holds fewer than 150 lifters the tool falls back to all ages and says so, which beats quoting a percentile off a handful of results.
What a percentile is good for
Calibration, mostly. It answers whether a number is normal for someone your size, which helps when you are working out if a lift is a strength or a weak point relative to your others. A lifter at the 70th percentile on deadlift and the 30th on bench has learned something they can act on.
It works less well as a target. Percentiles move when the population moves, and chasing one is a good way to train for the table instead of the sport. Programming that works moves your own numbers over a block. Where that leaves you against strangers is a byproduct.
Using the raw data
The percentile table behind this page is available as a JSON file, every cell with its sample size. OpenPowerlifting's own data is in the public domain and downloadable as bulk CSV if you want to compute something different.
If you coach, the reason to look at a distribution is usually to decide what to program next. Cleanpull is where the block after that decision gets built: one percentage progression, personalized to each athlete's current maxes.
Common questions
My percentile looks low. Am I weak?
Probably not. You are being measured against people who compete in powerlifting, which is a small and self-selected slice of everyone who lifts. A result in the bottom third of this field would still be strong in most gyms. The page answers where you would place at a meet.
Why is there no beginner or intermediate label?
Because those labels would be wrong here. Every lifter in this dataset has competed in a sanctioned meet, so calling the bottom of that field "beginner" misdescribes both the data and the person reading it. The bands are quintiles of a competitive population and they are named accordingly.
My gym numbers do not match these. Why?
Meet lifts are judged: depth on the squat, a pause and a press command on the bench, a clean lockout on the deadlift. Most gym maxes never face that standard, and the gap commonly runs five to fifteen percent. Compare judged numbers to judged numbers.
Does this cover the snatch and clean & jerk?
No. OpenPowerlifting records powerlifting meets, so it covers the squat, bench press, and deadlift. Olympic weightlifting results live in separate federation databases with much smaller and more elite samples, which would make for a considerably less useful comparison.
How current is the data?
The snapshot date sits under the table. OpenPowerlifting updates continuously, and this page is rebuilt against a dated export instead of querying live, so the numbers hold steady between refreshes.
Is my data sent anywhere?
No. The whole dataset ships inside the page and every calculation runs in your browser. Nothing you type leaves the device.
Coach more.
Spreadsheet less.
Cleanpull turns one percentage-based block into every athlete's working weights, then keeps the logs and lift videos in one review queue. Every workspace starts free.