No account · no email · nothing stored
How much should
you bench?
Not a tier somebody invented. Where your lift would actually place, measured against 467,914 lifters judged at drug-tested raw meets.
Competing category, because that is how the classes are drawn.
Your best single
100 kg bench press · 83 kg men
Ahead of 20% of the class
Measured against 66,947 lifters whose best bench press was passed by three judges at a drug-tested meet, raw, in this class. Not a gym average: these are people who entered a competition.
79.4 kg50th
120 kg95th
162.5 kg
100 kg sits at about the 20th percentile of this class.
Next rung: 105 kg is 5 kg away and would put you ahead of 25% of the class.
Copies the number, the class and the population it was measured against.
Where these numbers come from
Judged lifts, not logged ones
Every other standards page in this category is built on numbers people typed into a training app about themselves. That is selection-biased twice over: people who log are stronger than people who do not, and nobody checked any of it. This is the other thing.
OpenPowerlifting
Snapshot 2026-09-05 · 4,030,219 competition entries
An archive of results from sanctioned powerlifting meets. Licence read on 2026-09-10 at the project's own repository: all CSV data is waived of copyright and neighbouring rights, and is contributed to the public domain. Attribution is explicitly not required. It is rendered below anyway.
Raw, and drug-tested
1,421,028 entries kept · 467,914 lifters
No suits, no shirts, no wraps, and only federations that test. Both filters shrink the sample and both make the comparison mean something for somebody standing in a normal gym asking a normal question.
One row per lifter
Best result per lifter, per lift
Using all four million entries would weight the table towards whoever competes most often, which measures how much somebody enjoys competing rather than how strong the field is. Each lifter appears once per lift, at their best, in the class they weighed in at that day.
The percentile itself
Nearest-rank at fourteen breakpoints, linear between them
Fourteen points rather than four hundred, because the thin classes cannot support that many claims. Between two real points the line is straight, and outside the range the page stops rather than extrapolating. No curve was fitted and no distribution was assumed.
This page uses data from the OpenPowerlifting project, https://www.openpowerlifting.org. You may download a copy of the data at https://gitlab.com/openpowerlifting/opl-data.
Questions
- Is this how strong I should be, or how strong competitors are?
- Competitors. Everybody in this table entered a meet, weighed in, and had the lift passed by three judges. People who compete are much stronger than people who do not, so the fiftieth percentile here is nowhere near a gym average. That is why the answer is phrased as where you would place rather than as a tier of human you belong to.
- Why only drug-tested meets?
- Because an untested benchmark is the wrong answer to the question almost everybody is asking. Filtering to drug-tested federations costs sample size and gives a comparison that means something for a natural lifter. It is not a guarantee, it is a testing regime, and there is a difference.
- Why raw only?
- Raw means no squat suit, no bench shirt, no knee wraps. If you are asking what you should bench, you are not asking in a bench shirt, and equipped lifts add enough that mixing them in would make the table answer a different question.
- Where do the weight classes come from?
- They are the IPF classes the sport actually uses: 59, 66, 74, 83, 93, 105, 120 and over for men, 47, 52, 57, 63, 69, 76, 84 and over for women. The edge belongs to the lighter class, so 83.0 kg is an 83 and 83.1 kg is a 93. Comparing you against the people you would be called to the platform with is more meaningful than comparing you against everyone within five kilos.
- Why does it refuse to answer sometimes?
- Two reasons, and both are deliberate. A class with fewer than 300 lifters is left out, because a percentile that moves by whole kilograms when one person competes is not a percentile. And past the top or bottom of the table it says so instead of extrapolating, because the shape of the tail is not in the data and inventing it would undo the point of the page.
- Do you store any of this?
- No. The table ships with the page, the arithmetic runs in your browser, and nothing is sent anywhere. There is no account, no email box and no cookie behind this page.
A percentile is not a target. Nothing on this page knows your training age, your injuries or what you are trying to do. Being behind the class you would compete in is the normal state of almost everybody who lifts, including most people who eventually place in it.
One number is a snapshot
The useful question is not where you place today, it is whether the line is going up. Spottr works your one-rep max out of every set you log, so you can watch it move instead of re-typing it into a page like this one every few months.