“Deadlifts send testosterone through the roof” is one of lifting’s most durable beliefs. The real story is better: the spike is small, it’s over within the hour, part of it isn’t even new testosterone — and when researchers tested whether it builds muscle at all, it didn’t.
The claim has a satisfying logic to it: deadlifts are the heaviest, most whole-body lift you can do, so surely they flood you with testosterone, and surely that’s why they build muscle. Every part of that chain is worth checking, because when you check it, almost none of it survives. Heavy compound lifting does produce a real, measurable testosterone rise — that part is true. But it’s far smaller than advertised, it’s gone within about an hour, a good chunk of it is a measurement artifact rather than new hormone, and — the decisive part — it has been directly tested as a driver of muscle growth and came up empty. This isn’t a technicality. It removes one of the most common reasons people force themselves through lifts their body isn’t ready for.
Here are the real numbers. After heavy compound training, testosterone rises roughly 10–30% above baseline, peaks around 15–30 minutes post-set, and is back to baseline within about an hour. That’s the whole event. Set against a hormone that naturally swings substantially from morning to evening, the training “spike” is smaller than your own daily variation. It isn’t through the roof; it’s a blip inside the noise.
This is where it gets genuinely interesting. Three separate things drive that measured number up, and only one of them is actually more hormone.
Hemoconcentration. Hard training pushes plasma water out of the bloodstream and into the working muscle. Less water, same amount of testosterone — so the concentration in the blood draw rises. Part of the famous spike is literally the measurement of a more concentrated sample. Nothing was produced.
Reduced hepatic clearance. During heavy sets, blood is shunted toward working muscle and away from the liver — which is the organ that clears testosterone. Less clearance means it accumulates transiently. Again: not more production, just slower removal.
Genuine Leydig stimulation. Catecholamines and lactate do signal the testes to produce somewhat more. This part is real new hormone — and it is the smallest of the three contributors.
The magnitude tracks muscle mass recruited × load × volume × short rest periods. That’s genuinely why deadlifts and squats headline the studies — they recruit more muscle under more load than almost anything else. The mechanism is real. It’s the size and the consequence that were oversold.
Suppose we grant the spike entirely. The chain still breaks at the second link — and this was tested head-on. The well-known work from West and Phillips compared training conditions engineered to produce large acute hormone spikes against conditions producing minimal ones, then measured what actually matters: muscle protein synthesis and hypertrophy. The result: no meaningful difference. The acute hormonal response turned out to be a byproduct of training hard, not a cause of the adaptation.
What drives growth instead is local: mechanical tension in the trained muscle and the mTOR signaling it triggers, right there in the tissue you loaded. Not an hour of slightly elevated systemic hormone. This is also why you can’t “deadlift for biceps” — if a systemic hormonal wash were the mechanism, every muscle would grow from leg day. They don’t.
None of them meaningfully. Squats, Olympic lifts, high-volume leg press, and short-rest full-body circuits all produce a comparable bump — which is to say, the same small transient blip. There is no lift with a hormonal trick the others lack, so there’s nothing here to chase. It’s worth noting the opposite risk too: chronic excessive volume with poor recovery actually lowers testosterone, by driving cortisol up and suppressing the HPG axis. So “train more for T” isn’t merely useless — pushed far enough, it inverts.
The unglamorous list — roughly in order of leverage:
One of the most common reasons people push through pain or instability on heavy deadlifts and squats is precisely this belief — that these lifts hold unique hormonal magic nothing else offers. They don’t. Which means the deadlift and the squat have no hormonal claim on you. If a movement isn’t ready for you — or you for it — there is no testosterone argument for forcing it anyway. Choosing the lift your body can actually do well costs you nothing hormonally, and that’s one less reason to override your own signals.
Heavy lifting raises testosterone by only 10–30% for under an hour — less than your normal daily swing — and a large part of that “rise” isn’t new hormone at all, just hemoconcentration (thicker blood, same testosterone) and reduced liver clearance (less removal), with genuine production the smallest slice. And even the real spike doesn’t build muscle: when tested directly, high-spike and low-spike training produced the same hypertrophy — growth is driven by local mechanical tension and mTOR, not systemic hormone. No lift boosts T meaningfully, and chronic overtraining lowers it. The real levers are body fat, sleep, recovery, and correcting deficiencies — so no lift has a hormonal claim on you.
This article is for educational purposes only and is not medical advice, diagnosis, or treatment. Hormonal physiology varies between individuals, and low testosterone can have medical causes that deserve proper evaluation — symptoms of concern should be assessed by a physician with appropriate bloodwork rather than addressed through training changes or supplements. Figures cited are representative of published findings in trained and untrained populations and are not guarantees of individual response. Exercise carries injury risk; introduce load gradually and with attention to form, and anyone with a musculoskeletal issue should be assessed in person by a qualified clinician. Nothing here recommends a specific program, exercise, load, or intervention for any individual.
This lesson relates to these health systems — health works as a connected system, not isolated topics.
Prerequisite: Two Hands on One Lever: Stacking Agonists on One Axis