Semen retention: the 7-day testosterone spike is real, and it's a trap

The famous number, and what happened to it

The study everyone quotes, Jiang et al., Journal of Zhejiang University, 2003, followed 28 volunteers who ejaculated once at baseline and then abstained while blood draws tracked testosterone daily. From day two to five, minimal fluctuation. On day seven, a clear peak: 145.7% of baseline. After the peak, no regular pattern. That single peak, from a single small uncontrolled study, became the retention movement's founding scripture, quoted in thousands of videos as if it were settled physiology.

It isn't settled. The journal retracted the paper in December 2021 because it substantially duplicated an earlier Chinese-language article by the same lead author, a retraction for redundancy and duplicate publication. Retraction doesn't by itself prove the data wrong, and to be fair to the movement, nobody has demonstrated fraud either. But it means the number has been pulled out of the record by the journal that published it, and everything that was built on it was built on sand. Even before the retraction, the study had the problems a viral claim can't afford: 28 men, no control group, no replication, and a result that, by its own admission, went nowhere after day seven.

So the "45% spike" is best described honestly: one small study once found a day-seven peak, the study has been retracted, nobody has reproduced the finding since, and the authors themselves reported no sustained elevation afterward. That is not a superpower. That is a footnote that went viral.

What ejaculation actually does to your hormones

The short version: a transient, self-correcting shuffle, not a withdrawal from your hormonal bank account.

The endocrine literature on orgasm is reasonably consistent. During and right after orgasm, prolactin rises and dopamine dips, which is the biology of the refractory period, the satisfied, briefly disinterested state that follows. Testosterone itself doesn't meaningfully move with orgasm in the studies that measured it directly. What happens next is the part the retention movement gets wrong: the small prolactin rise briefly nudges luteinizing hormone down, LH being the pituitary's signal to the testes to make testosterone, and levels can drift modestly below baseline for a window measured in hours, not days. Then the axis resets. There is no cumulative drain, no depletion, no compounding loss across ejaculations. The system is a thermostat, not a tank.

One careful study of hormonal responses to orgasm, with and without a preceding abstinence period, is worth naming because it settles the argument cleanly: orgasm produced its usual endocrine response either way, and abstinence didn't change how orgasm behaved. Abstinence did raise basal testosterone modestly in that study, echoing the day-seven idea at smaller scale, but a modest, transient rise in a baseline reading is not the same thing as more androgenic effect, and nothing in the literature shows a retained semen regimen translating into more muscle, more strength, or more drive than a normal sex life. Those claims are the movement's, not the evidence's.

The finding that points the other way

If ejaculation frequency matters for men's health, the best data we have says frequent is better, and the outcome it measured is not a hormone blip but cancer. The Health Professionals Follow-up Study tracked ejaculation frequency in tens of thousands of men over decades. In the 2016 update, men reporting 21 or more ejaculations per month in middle age had roughly a third lower risk of being diagnosed with prostate cancer than men reporting four to seven per month, and a later pooled analysis put the difference at around 50% for the highest-frequency group. It's observational data, so it shows association, not proof of causation, and it has to be read with that caveat held tight. But it is large, prospective, repeated across populations, and it points squarely against the idea that holding is protective.

Read the two bodies of evidence side by side and the asymmetry is stark. For retention: one retracted study of 28 men showing a transient peak, plus testimonials. For normal-to-frequent ejaculation: stable hormones in direct measurements, plus the largest cohort data available pointing at a prostate-cancer association in the favorable direction. Whatever else retention is, it is not the evidence-backed position.

What retention might still be good for, honestly

None of this means every day-counting habit is worthless, and the honest version of this article has to say what retention plausibly does deliver. If a man's sexual life is running on compulsive pornography use that's interfering with his sleep, his relationships, or his response to real partners, changing that pattern can help, but the mechanism is behavioral, not hormonal: it's the recalibration covered in the death-grip piece, and large cross-national studies find no association between pornography use as such and sexual dysfunction, which suggests the benefit, when there is one, comes from changing a compulsive pattern, not from retaining semen. The discipline, focus, and self-efficacy that retention communities report are real experiences; they just don't require a hormonal explanation, and they don't require the retracted one.

So the fair synthesis: the movement's hormone claims don't survive contact with the literature, but the self-regulation framework can be a legitimate tool for men working on compulsive patterns, and it works whether or not anyone holds for seven days.

The bottom line

The 7-day spike is real as a historical claim, and a trap as a strategy: one small study, retracted, never replicated, describing a blip that even its authors didn't show persisting, sold for two decades as the key to manhood. The hormones don't drain, the thermostat resets, and the largest data we have on the question associates frequency with lower prostate-cancer risk, not higher. If your goal is discipline, keep the structure and drop the endocrinology. If your goal is hormones, this isn't the lever, and reading your actual labs is where that conversation starts.

Common questions