Everyone in the gym knows how deep they squat. Almost nobody is right.
This is not a character failing, it is a perception problem, and it is one of the most reliably reproduced observations in the entire folk literature of lifting: ask a lifter to squat to parallel and they will consistently stop above it, and be certain they did not. The bottom of a squat is the single worst moment in your training week to be estimating anything. You are upside down, under load, and your proprioception has other commitments.
So the useful version of this article is not "squat deeper." It is: stop guessing, and film it.
The standard, and what it is actually for
The common definition: hip crease below the top of the knee. Below parallel. It comes from powerlifting, where somebody has to judge lifts and a shared line was required, and it has leaked into general training as a kind of moral threshold, which was never the intention.
For most people, most of the time, below parallel is a good target. Full-range work through a deep squat asks more of the glutes and adductors, trains the position you are trying to keep, and gives you a consistent bottom, which matters enormously for progression, because a squat that gets three centimetres shallower every week is not a squat that is getting stronger. It is a squat that is getting shorter, and the log cannot tell the difference unless something measured it.
But depth is conditional. Ankle dorsiflexion, hip anatomy, femur length, and what your lower back does at the bottom all vote. If your pelvis tucks under and the spine rounds (the butt wink, the one thing in this article with a genuinely silly name), you have found the end of your usable range, and the reps below it are being paid for by tissue that does not want them. That is where your depth is. Not where a stranger's is.
The rule that survives all the arguing: the deepest position you can reach and control, and reach the same way every rep. Consistency is the part that shows up in the numbers.
The knee thing
Deep squatting does not destroy healthy knees. That belief has been circulating since the 1960s on the back of research that would not survive contact with a modern reviewer, and it will outlive us all.
What does hurt knees is the ordinary stuff: too much load, added too fast, with technique that degrades under it, on tissue that has not been given a reason to adapt. Depth is not the villain in that story; the rate of progression usually is.
If you already have a knee problem, this is a matter for someone who can examine you. I am a blog post. I cannot see your knee and would not know what to do with it if I could.
Why "just film it" is where this ends up
Depth is a geometry question. Geometry questions are answerable, which makes them unusual and precious in a field where most questions are not.
Set a phone on the floor to the side, film one working set, watch the bottom frame. Where is the hip crease relative to the knee? That is your answer, and it does not care how the set felt, which is exactly why it is worth having. Most people watch this footage for the first time and experience a quiet, private recalibration. Some of them squat deeper afterwards. All of them stop arguing.
Once you are filming, depth stops being the only thing you can see. Knee travel. Whether one hip rises before the other out of the hole, which is common, invisible from inside the movement, and by far the most useful thing a side-on camera will ever tell you. Tempo, and how much of the bottom position is being spent falling into it.
Doing it in OneRep
Form Coach takes a filmed squat and returns measurements rather than adjectives: joint angles through the rep, depth at the bottom, tempo, left-right symmetry. Then the corrections, with the numbers they came from attached, so you can disagree with it if you think it is wrong. That is the only honest way to ship this kind of feature.
Every account can film a set; a free one covers about five a month, and Coach raises the allowance. If you are curious how a rep counter with no table of exercises works out what it is looking at, we wrote that one up too, along with why the pose model got ten times slower when we optimised it.
The logger side is duller and matters as much: log the barbell full squat with last session's numbers sitting beside the set, and a depth cue in the notes if depth is the thing you are fixing this block. Consistent depth, consistent load, a trend you can read across twelve weeks.
Try it: Free, no trial window.
Browse: Exercise library · OneRep vs Strong