The highest-value fix on this list is not squatting more often — it's ending every set one or two reps short. In Pareja-Blanco's 8-week squat trial, the group that cut each set at 20% velocity loss gained the same squat strength as the group that ground to 40% velocity loss, while doing about 40% fewer reps. Frequency, which I led my own reel with, is a delivery system for volume rather than a lever on its own — three squat days help because they let you fit in more good sets, not because the calendar does anything by itself. So the practical version is short: squat two to three times a week, end every work set at 1-2 reps in reserve, and add 5 lb only when every work set hits the top of its rep range.
I made a reel with these seven fixes. This is the long version, with the studies attached — and with one of the seven walked back, because when I went looking for the evidence behind my own tip, it wasn't there. And if you're chasing a number instead of a program, three plates is the one almost everybody picks — here's whether 315 is a good squat, and why your bodyweight decides the answer.
How often should you squat to get stronger?
Two to three times a week — but the frequency itself isn't what makes you stronger. The extra volume it lets you fit in is.
Grgic's 2018 meta-analysis in Sports Medicine pooled 22 studies and found strength effect sizes rising with training frequency: 0.74 for once a week, 0.82 for twice, 0.93 for three times, 1.08 for four or more (p=0.003). Squat more days, get stronger.
Except in that same paper, the subgroup where total volume was matched showed no significant effect of frequency. And split by body region, the effect held for upper body (p=0.004) and not for lower body (p=0.070). Legs — the exact thing this article is about — is where it stopped reaching significance. That split is why frequency earns a higher spot on the list when the goal is to increase your overhead press than it does here. The authors also note most of the pooled studies used untrained participants.
Then there's the direct test. Colquhoun 2018 took 28 trained men through six weeks of powerlifting training spread over 3 versus 6 sessions a week, with volume and intensity equated. Squat 1RM rose about 17 kg in both groups (+16.8 vs +16.7 kg). No group-by-time interactions. Double the sessions, same result.
Here's the one I got wrong. My reel said frequency alone will add plates to the bar faster than any program switch. That's the line the research doesn't support, and it's least supported for legs specifically. Two to three squat days is still the right prescription — just for the honest reason: more days means more room for quality sets without one session turning into a two-hour grind. Same logic applies to every muscle group, not just the squat.
Why is my squat not going up?
Because you grind the last rep of every set, and that one rep quietly deletes reps from every set after it.
Pareja-Blanco 2017 put 22 young men through eight weeks of velocity-based squat training. One group stopped each set when bar speed dropped 20%; the other pushed to 40%. The 20% group got similar squat strength gains on 40% fewer reps, plus a bigger countermovement jump improvement (9.5% vs 3.5%, P < 0.05). The 40% group did get more vastus lateralis hypertrophy — so if size is the goal, the extra reps aren't wasted. For strength, they bought nothing.
The meta-analysis agrees, with a caveat worth quoting. Davies 2016 pooled eight studies on training to failure versus stopping short and found an effect size of 0.34 favouring non-failure (p=0.02), with compound exercises at 0.37-0.38. But the authors themselves call that difference "unlikely to be meaningful." Their conclusion isn't that failure hurts you. It's that failure is unnecessary.
Refalo 2024 shows the cost directly. Eighteen trained people did eight weeks of leg press and leg extension with one limb taken to failure and the other stopped at 1-2 RIR. Quadriceps thickness gains were similar, but velocity loss and rep loss within the session were consistently greater on the failure limb. That's the mechanism: the grinding rep doesn't come free, it comes out of the next set. Be fair to the paper, though — over the full eight weeks total volume came out similar between the two limbs. More on where to stop on a squat set.
What squat stance and depth should you actually use?
Pick the stance that lets you hit depth without your heels lifting or your back rounding — because depth is the part with real evidence behind it, and stance width is just how you get there.
Stance does change the job. Escamilla 2001 filmed 39 lifters at a national powerlifting championship and found a wide stance flexed the hips 6-11 degrees more, put the thighs 7-12 degrees more horizontal, and kept the shanks 5-9 degrees more vertical than a narrow stance. That's a redistribution of load between ankle, knee and hip. What it isn't is an upgrade — no study shows one stance width produces a bigger squat.
Depth is a different story. Pallarés 2020 split 53 trained men across 10 weeks of full, parallel, or half squats at matched relative loads, against a no-training control. The full-squat group was the only one to raise 1RM across all three squat variations (effect sizes 0.77-2.36). The half-squat group improved on none of them (-0.11 to 0.28) and was the only group reporting significant increases in pain, stiffness and functional disability. Kubo 2019 found full-squat 1RM rose 31.8% in the full group against 11.3% in the half group (p=0.003).
So this tip survives, but for the other reason. Set your stance wherever depth comes easy. Then go get the depth.
Do pause squats actually increase your squat?
Yes — this is the one tip with a direct 10-week trial behind it.
Martinez-Cava 2021 put 26 men through 10 weeks of identical velocity-based squat training. The only difference was a roughly 2-second pause between the eccentric and concentric versus a normal rebound. Both groups improved. The pause group posted effect sizes of 0.76-1.12 against 0.45-0.92 for rebound.
Be precise about what that means: those are larger within-group effect sizes, not a declared significant between-group difference on 1RM. "This is a strong result" is defensible. "Proven to add more to your 1RM" is not. And the rebound group wasn't along for the ride — countermovement jump and 10-20m sprint favoured them. Pausing trains the position; bouncing trains the spring.
Prescription: a 2-second dead stop at the bottom, 3-4 sets of 3-5 reps, at roughly 85-90% of your touch-and-go weight. Don't read that drop as a step backward — it's the same set with the elastic help removed.
How should you brace and control the way down under a heavy squat?
Big breath into the belly, brace like you're about to get hit, then take 2-3 seconds on the descent — the brace has a measured mechanism, the tempo mostly buys you position.
The brace is measurable. Blazek 2019 screened 1,125 papers down to a 16-study review and found the squat produces the highest intra-abdominal pressure of any exercise reviewed — over 200 mmHg, ahead of the deadlift and the leg press. That pressure is what stiffens the trunk under a loaded bar. Carry the authors' own note with it: if you're untrained or you have hypertension, progress into heavy Valsalva work gradually.
Here's the second one I got wrong. My tip #7 said a 2-3 second descent "builds more muscle" and "keeps your knees healthy." Neither holds up. Schoenfeld 2015 pooled eight studies and found hypertrophy was similar across rep durations from 0.5 to 8 seconds; only reps slower than about 10 seconds looked worse. And on the knee claim — there is no trial testing whether a controlled descent protects knees. I said it because it sounds true. That's not a reason.
The habit survives; my explanation for it doesn't. Control the descent so you arrive in the same bottom position every rep. That's what makes the load repeatable and the log honest — and the log is the only place you find out whether the last four weeks worked.
Does building your posterior chain carry over to a bigger squat?
Partly — hip thrusts do transfer to squat 1RM, but less than squatting does, so treat this as an addition to your squat work and not a substitute for it.
Hammond 2019 randomised 14 trained men to four weeks of back squats or barbell hip thrusts, twice a week at 80% 1RM. Both groups raised their back-squat 1RM, which settles the transfer question. But squatting won: +11.8 kg (10.2%) versus +8.6 kg (7.7%), p=0.050. Contreras 2017 found the same split of jobs over six weeks in adolescent male athletes — front squats better for front-squat 3RM and vertical jump, hip thrusts better for 10m and 20m sprint and the isometric mid-thigh pull.
And here's the third walk-back, this one softer. My reel said weak hamstrings and glutes equal a squat that falls forward. I can't point you to a study on that one — no trial tests it. The hamstrings work as hip extensors and knee flexors at once in a squat, which makes their net contribution messy, and forward lean at the sticking point is at least as much a quad, trunk-extensor and depth problem. Keep training the posterior chain — it builds bigger legs and the transfer is measured. Just don't let anyone sell it to you as the reason your chest is caving.
What does a week of squat training actually look like?
Two squat days you don't skip, plus optional light technique work — inside a 4-day split, not a squat-every-day program.
When I was 285 lbs I couldn't walk up stairs without getting winded. Now I squat 225 for 30 reps unbroken, and my 1RM is 425. That came from years of logged training on four days a week, with the weight going up in increments small enough to repeat. Taking that much weight off while the bar keeps going up is its own problem with its own levers — that's how to lose fat without losing muscle.
Day 1 — heavy
3-4 sets of 3-5 reps at 1-2 RIR. This is the day the number on the bar moves. If the last rep of set one is a grind, the weight was wrong, not your effort.
Day 2 — pause
3-4 sets of 3-5 with a 2-second dead stop, at roughly 85-90% of your Day 1 weight. Same rep target, different job: this one owns the bottom position.
Accessories
Added to your squat work, never instead of it:
- Romanian deadlift — 3 sets of 6-10
- Leg curl — 3 sets of 8-12
- Hip thrust — 3 sets of 6-10
The progression rule: when every work set hits the top of its rep range at 1-2 RIR, add 5 lb. Not when you feel strong, not when the weight looks light — when the log says every set cleared the range. That one rule is what turns two sessions a week into a bigger squat over months.
If you want the whole week around it, this is the 4-day split and the one-screen scorecard — written as a day-one starting layout rather than my personal program, with the same overload rule attached.