Search for semen retention benefits and you will find lists: more energy, sharper focus, higher testosterone, more confidence, better training sessions, a kind of magnetism other people supposedly notice. You will also find a great deal of confidence about why. This page goes through those claims one at a time and states what the evidence for each actually is, including the places where it points the other way and the places where there is none.
The short version before the detail. The experiences men describe are real descriptions: something is being noticed. The mechanisms offered to explain them are, with few exceptions, not established. And the reason this is unresolved is not that nobody has bothered. It is that no controlled trial of semen retention appears to exist in the peer-reviewed literature, that nearly all published abstinence research studies pornography rather than ejaculation, and that a man who takes up retention almost never takes up only retention.
What this page assesses, and what it does not
Four commitments get treated as one thing and are not. NoFap generally means abstaining from pornography and masturbation, with the original emphasis on porn. Semen retention means avoiding ejaculation by any route, including partnered sex. Hard mode usually means no porn, no masturbation and no orgasm of any kind. Celibacy means abstaining from partnered sex and says nothing about masturbation. A man can be doing hard mode and not be celibate; he can be celibate and masturbate daily. This matters constantly below, because a study of seven-day pornography abstinence is not a study of ejaculation abstinence, and forum posts trade them as if they were.
Non-ejaculatory orgasm is a fifth thing again, a technique rather than a commitment. It is documented: an interview study of 21 men who were already multiply orgasmic reported that detumescence does not always follow orgasm and that a series of orgasms is possible. That establishes the phenomenon occurs in some men. It says nothing about how common it is, how reliably it can be learned, or whether practising it delivers anything, none of which has been studied.
Practices of continence also sit inside long religious and philosophical traditions: brahmacharya in Hindu and Jain thought, the conservation of jing in Daoist practice, Ayurvedic accounts of shukra and ojas, and the nineteenth-century Anglo-American purity and anti-masturbation movements. These are distinct systems with distinct reasoning, not one idea in four costumes, and each argues from its own premises about discipline, vitality and the ordering of a life. None of them is making the empirical claims tested below, none of them is an early draft of a wellness protocol, and this page does not assess them. What follows is about modern claims stated in modern empirical terms.
Five verdict labels are used below, and they mean specific things.
- Supported: the specific thing claimed has been measured, more than once, and the measurements agree.
- Mixed: real evidence exists on both sides, or the answer depends on which outcome is meant.
- Unsupported: nothing establishes the claim, and the evidence usually offered for it does not bear on it.
- Contradicted: the thing has been looked for and what was found runs the other way.
- Untested: nobody has measured it. This is the most common answer on this page.
Testosterone, the mechanism most of the claims run through
Almost every benefit on the standard list is explained by a testosterone rise. So the chain has two links: does abstinence raise testosterone, and does raised testosterone deliver energy, confidence and drive. Both links need to hold.
Testosterone peaks on day seven at 145.7% of baseline, which is why the seven-day mark matters.
Jiang and colleagues measured serum testosterone daily in 28 volunteers during abstinence and reported minimal movement on days 2 to 5, then a peak on day 7 at 145.7% of baseline. The English-language paper everyone cites was retracted in 2021 for substantial overlap with the same group's earlier Chinese-language publication, not for fabricated data. It has never been independently replicated in over twenty years, had no control group, and measured no energy, mood, focus or performance outcome of any kind. It cannot carry the claim, and it is not evidence against it either. It is a small withdrawn study that measured nothing anyone practising retention wants to know, and it is discussed here as an object of study rather than used as a source in either direction.
Testosterone keeps climbing the longer you retain, so longer streaks mean higher levels.
No adequately powered longitudinal study has tracked serial morning testosterone in men across abstinence of a month or more. Not one, in either direction. For a claim this heavily marketed and this cheap to test, that gap is itself worth noticing. The charts circulating with a line rising through week one and onward for months have no dataset behind them: the continuation is drawn, not measured.
Anyone stating a number for day 30, day 60 or day 90 is not reporting a finding, because no finding exists.
Three weeks of abstinence raises testosterone.
Exton and colleagues studied 10 healthy men: a laboratory arousal and orgasm session, three weeks of abstinence, then the same session again. Orgasm itself did not alter plasma testosterone. Higher testosterone was observed after the abstinence period. It is a prospective within-subject human design with a longer window than anything else in the area, which is worth something. It is also 10 men, with no control group and a fixed session order, so any three-week drift from season, sleep, training or familiarity with the procedure is fully confounded with abstinence. It has not been replicated in 25 years, and secondary sources disagree about whether the difference appeared at resting baseline or only during arousal.
Ejaculating crashes your testosterone.
Several studies have looked for the drop and not found it. Exton measured plasma testosterone through masturbation-induced orgasm and found it unaltered. Isenmann's randomised crossover, 11 men enrolled and 8 completing, found masturbation and visual stimulus both appeared to counteract the normal daily decline in free testosterone, so the sexual condition ended higher. In a British national probability sample of 1,599 men using mass-spectrometry salivary assays, higher testosterone went with more recent and more frequent masturbation, not less. None of these are large definitive trials and the national sample is cross-sectional, so causal direction is open. But the drop the claim requires has been looked for and has not been observed.
Does more testosterone buy energy, confidence and drive?
More testosterone means more energy, more confidence and more drive.
The Testosterone Trials randomised 790 men aged 65 and over with testosterone below 275 ng/dL to a year of testosterone gel or placebo. The Sexual Function Trial (459 men) met its primary outcome, with moderate increases in sexual activity and desire. The Vitality Trial (474 men, selected for low self-reported vitality) missed its primary outcome, though the same paper reports small improvements in vitality, mood and depressive symptoms on the continuous measures. The Physical Function Trial (387 men) missed its primary endpoint, while walking distance improved significantly across all participants. Graded-dose work separately shows lean mass and strength do rise with testosterone dose across and above the physiological range, so dose-response is real for body composition. For the outcomes on the retention list it looks more like a threshold: in the same 1,599-man national sample, salivary testosterone showed no clear association with sexual function, erectile difficulty or lack of interest; and in 41 men giving daily saliva samples and daily desire ratings across 31 consecutive days, 759 observations, a man's own testosterone did not positively predict his own desire.
So the first link is thin and the second is conditional. Correcting a genuine deficiency does some things. Moving within the normal range is not shown to do the things practitioners are chasing, and abstinence has not been shown to move the number durably in the first place. A weak link chained to a conditional one is not a mechanism.
Energy, focus and mental clarity
Retention gives you more energy and clearer focus, and you feel it within a week or two.
No study has measured energy, focus, attention or cognitive performance in men practising deliberate ejaculatory abstinence. Not over a week, not over ninety days. The hormone studies that exist measured hormones and nothing else: neither Jiang nor Exton recorded a single subjective or behavioural outcome. This is the largest gap on the page. The reports are numerous and the measurements do not exist.
Abstaining resets or upregulates your dopamine receptors, and ninety days is a full reboot.
No human study has measured dopamine receptor availability, dopamine release or dopaminergic tone before and after a period of sexual abstinence, by PET or any other method. The claim is imported from rodent work on sexual satiety and from human addiction imaging that did not involve sexual behaviour. The recovery timelines quoted, fourteen days, ninety days, have no human measurement behind them. The related dopamine detox idea is a repackaging of stimulus control, a standard behavioural technique; stimulus control works, and the neuroscience wrapped around it is decoration. Dopamine is not a tank that empties and refills over a weekend.
Confidence, attraction and magnetism
After enough retention people notice you: more magnetism, more attention, women approaching.
No study has tested whether abstinence changes how others perceive or respond to someone. What is available runs the other way as explanation rather than refutation: a man who has committed to a difficult practice, counts it daily and expects to become magnetic will notice attention that was already there. Expecting to be more confident often does produce more confident behaviour, and that genuinely changes how people respond, which makes the experience real while leaving the causal story unestablished. In a survey of 1,063 respondents recruited through a non-thematic Reddit thread, abstinence motivation tracked beliefs, chiefly the belief that masturbation is unhealthy, along with conservatism, religiosity and lower trust in science, rather than any behavioural marker. That is a population with strong priors about what should happen.
Gym performance
Do not ejaculate before a heavy session or a competition: more testosterone means more muscle.
The gym version chains two claims. The first, that abstinence raises testosterone, rests on the withdrawn study above and on nothing that has replaced it. The second, that a systemic testosterone rise builds muscle, was tested directly in 49 resistance-trained men over 12 weeks of training: hypertrophy and strength improved similarly across conditions, with no significant correlations between the acute post-exercise rise in systemic hormones and any outcome. That removes the acute-spike mechanism. It does not test chronic baseline testosterone in abstaining men, which nobody has done. On competition specifically, a systematic review found the evidence does not support a general performance decrement from sexual activity, with the interval mattering most and a negative effect appearing only under roughly two hours. Nothing shows that extended abstinence improves athletic performance.
The clearest counter-evidence: ejaculation frequency and prostate cancer
In the Health Professionals Follow-up Study, 31,925 men reported their average monthly ejaculation frequency in 1992 and were followed to 2010: 480,831 person-years, 3,839 prostate cancers. Men reporting 21 or more ejaculations per month had lower incidence than men reporting 4 to 7, with a hazard ratio of 0.81 for frequency at ages 20-29 and 0.78 at ages 40-49. A 2025 meta-analysis of 29 studies and 315,193 participants found a pooled odds ratio of 0.83. The authors of the cohort tested the obvious objection, that this is an artefact of PSA screening, and the association held within the screened subgroup.
Retaining semen is good for your prostate.
The one large, repeatedly observed association in this area runs in the opposite direction to the claim. It concerns ejaculation frequency rather than deliberate retention, and no study has ever measured a retention group, so it is not proof that retention causes harm and should not be read as one. It does mean the benefit claim has the best available observational evidence against it rather than for it.
Where that finding is soft, and where it is not
The limits are worth stating precisely rather than as a general shrug. The cohort is observational, the exposure is self-reported recall of frequency decades earlier, and the participants are US male health professionals. The 2025 meta-analysis pools case-control with cohort designs and includes the same cohort analyses, so it is not independent replication; the same analysis found masturbation frequency specifically was not significantly associated with risk, which sits awkwardly alongside the ejaculation-frequency result. The association is concentrated in low-risk, screen-detected disease. Whether frequency relates to advanced disease or to death from prostate cancer is unresolved: the earlier analysis of the same cohort had only 147 advanced cases, too few to demonstrate either an effect or its absence. And a case-control study of 431 cases and 409 controls diagnosed at 60 or younger found the opposite direction for activity in the twenties and thirties. Those are genuine inconsistencies in the evidence base. They do not convert the finding into support for the claim.
What measurably changes in semen
Holding it in for weeks builds up better, stronger semen.
It depends entirely on the parameter, and the parameters move in opposite directions. In an analysis of 9,489 samples from 6,008 patients, mean volume rose from 2.3 mL at 0 to 1 days of abstinence to 3.9 mL at 8 to 14 days. In the same dataset, peak motility in oligozoospermic samples came after one day, peak normal morphology between 0 and 2 days, and normozoospermic samples showed significant falls in motility and normal morphology by days 11 to 14; the authors recommend not exceeding 10 days. A 2024 meta-analysis of 13 studies and 2,315 men found longer abstinence gave higher concentration and volume but lower progressive motility. More fluid, containing more sperm, that swim less well. A 2026 literature review of the fertility evidence concludes that shorter intervals are associated with better sperm function and that routine abstinence beyond three to five days appears unnecessary and possibly counterproductive for men attempting conception; it is a narrative review framed around fertility treatment rather than general health. All of this comes from fertility-clinic populations, and none of it was measured for any outcome retention practitioners care about.
Longer abstinence is associated with higher sperm DNA fragmentation on laboratory assays.
The 2024 meta-analysis found fragmentation significantly higher after longer abstinence, with a meta-regression showing days of abstinence positively and linearly related to it, slope 0.65, p=0.044. A 2022 systematic review found that in men with abnormal parameters, a second ejaculate within four hours had a significantly lower fragmentation rate. The effect size is modest, the assays differ between studies, and the clinical meaning of a given fragmentation percentage is itself contested within andrology. This is a measurable change in a laboratory parameter across days. It is not a demonstrated harm, and the relationship was fitted over a range of a few days, so extending it to weeks or months is unwarranted.
The depletion model, checked with arithmetic
Every ejaculation costs you a serious amount of zinc, so retention preserves your zinc status.
This one does not need a trial. Measured total zinc per ejaculate in 152 men averaged around 305 to 336 micrograms depending on the subgroup, across a range of roughly 220 to 440. The recommended dietary allowance for adult men is 11 mg per day and the tolerable upper intake level is 40 mg. One ejaculate is therefore about 0.3 mg, roughly 3% of a single day's recommended intake; a full month of daily ejaculation totals less than one day's allowance. The figures of 5 mg or 9 mg per ejaculate that circulate appear to come from reading a per-litre concentration as a total content. Zinc in seminal plasma has real functions and is measured clinically as a marker of prostate secretory function. The quantity lost is not nutritionally meaningful for a man eating an ordinary diet.
Total zinc per ejaculate averaged 336 micrograms in normal-viscosity samples across 152 men (Dissanayake 2010). Reference values from the Institute of Medicine dietary reference intakes. The first bar is barely visible, which is the finding.
See the numbers
| One ejaculate | 0.336mg |
|---|---|
| Daily RDA, adult men | 11mg |
| Tolerable upper intake | 40mg |
Erections and morning erections
Retention improves erectile function and brings morning erections back.
Nobody has measured erectile function in men practising deliberate retention. The only prospective population data on the adjacent question point the other way, in a much older group and for intercourse rather than ejaculation: in a Finnish five-year follow-up of 989 men aged 55 to 75 who were free of erectile dysfunction at entry, men reporting intercourse less than once per week had roughly twice the incidence of moderate or complete dysfunction, an adjusted incidence rate ratio of 2.2. Reverse causation is a serious problem there, since early difficulty reduces frequency before it is reported as dysfunction, and the men are decades older than typical practitioners. Worth noting separately: that study found no relationship between morning erections and later erectile dysfunction, and morning erections are one of the most commonly tracked self-reported markers in these communities.
The flatline
Around week two or three you hit the flatline, and pushing through it means the brain is rewiring.
The most discussed phenomenon in these communities has essentially no peer-reviewed literature behind it, and NoFap's own FAQ says so, adding that there is no way to predict when or whether it occurs, with reports ranging from days to months. The nearest test is a randomised controlled study of seven-day pornography abstinence: 176 undergraduates, 86 abstaining and 90 controls, in a Malaysian sample that was 64.2% female. It found no significant group differences in craving, affect or withdrawal symptoms, and 45.35% of the abstinence group lapsed at least once inside the week. Seven days, pornography rather than ejaculation, and mostly women: it cannot speak to a flatline claimed to begin at week three. The claim has not been examined.
One thing does need saying plainly. The reported flatline picture, low libido, flat mood, anhedonia, low motivation, fatigue, overlaps substantially with the clinical picture of depression, and nothing in the community vocabulary distinguishes between them. Persistent low mood, loss of motivation that does not lift, pain, or erectile problems that do not resolve are reasons to see a doctor, not stages in a protocol to push through. NoFap's own guidance says the same. Treating a depressive episode as a milestone is the most serious risk in this discourse.
Why population studies probably will not settle this
A man who starts retention almost never starts only retention. The same week typically brings resistance training, an earlier bedtime, meditation, cold exposure, less alcohol, deleted apps and a new social identity inside a community that expects improvement. Every one of those has independent, well-documented effects on mood, sleep, energy and self-rated confidence. When he feels better at day thirty, the observed effect is the sum of all of them plus expectancy plus the season, and retention is one term in a long sum. A between-subjects study makes it worse rather than better, because men who adopt retention differ systematically from men who do not. Randomising is the standard fix, and you cannot blind a man to whether he is ejaculating, so expectancy stays in the design permanently.
You cannot blind a man to whether he is ejaculating, so expectancy stays in the design permanently.
There is a further problem with self-report from this population. A systematic review and meta-analysis found that moral incongruence about pornography use, the experience of acting against your own beliefs, was associated with greater distress about use, greater general psychological distress, more reported problems and higher perceived addiction, independent of how much a man actually used. A question like "is this out of control" is therefore partly measuring what a man believes about the behaviour rather than the behaviour itself. Two men doing identical things return different answers, which is one more reason a between-person average is hard to interpret here, and one more reason a man's own baseline is the only comparison that holds still.
The improvements men report on retention are effects of retention.
They may well be real experiences. Attributing them to retention specifically is the step the evidence does not license, and three named mechanisms would produce identical reports even if retention did nothing. Expectancy: the reported benefits are almost all subjectively rated, and there is no blindable version of this. Measurement reactivity: a systematic review of 19 studies found consistent evidence that being observed or asked to report on your own behaviour changes that behaviour, though the size was inconsistent. Monitoring itself: a meta-analysis of 138 randomised studies with 19,951 participants found that prompting people to monitor progress toward a goal reliably improved performance and goal attainment, with larger effects when progress was recorded. That last one cuts both ways and should be said plainly: tracking probably helps, and tracking is therefore also a confound. On top sits survivorship, since non-responders quietly stop posting.
Thirty days is the first real milestone.
The most directly relevant prospective data come from a study of No Nut November participation: 435 respondents before the month and 114 after, roughly 74% attrition, self-selected and self-reported throughout. Participants reported greater sexual flexibility than people who had never taken part, but across the abstinence period no measures of sexual wellbeing differed significantly between participants and non-participants. The authors conclude a month of abstinence may be associated with neither improved nor worsened sexual wellbeing. With that attrition it is a weak null, and it measures sexual wellbeing rather than mood, focus or energy. It is still the only prospective look at a thirty-day mark, and it found nothing at it.
The streak, and what resetting it does
The all-or-nothing relapse framing used in these communities is harmless.
A preregistered survey of 417 men who had heard of Reboot, 257 of whom had attempted it, found greater involvement with NoFap forums associated with more reported current depression, anxiety and erectile difficulty, and described the most recent relapse as commonly followed by shame, worthlessness and sadness, with 28.9% of those who relapsed reporting suicidal feelings. Respondents also reported observing trolling (88%), misogyny (73.7%) and bullying (49.1%), with 23.5% reporting being told to harm themselves. This is cross-sectional and correlational: the article itself says the direction is unclear and that people with more mental health symptoms may be likelier to seek out these communities. It was recruited by online advertisement, and the lead author is a prominent public critic of NoFap, which is worth stating openly rather than using to dismiss or inflate the result. What it does not support is the assumption that the framing is neutral.
The question that does have an answer
Notice what the honest verdicts above have in common. Where evidence exists it is between-person: an average across thousands of men, from which nothing follows about any particular one. Where the practice itself is concerned, the evidence mostly does not exist, and the confounds are not the kind that a bigger sample fixes. "What are the benefits of semen retention" is a question the population literature is not in a position to answer, and given the design problems it may never be.
There is a smaller question that is answerable. Within one person, holding most things roughly constant, changing one variable at a time and measuring the outcome repeatedly is a recognised methodology: the n-of-1 trial, formalised by the Agency for Healthcare Research and Quality as an approach to single-patient comparative effectiveness. It does not escape expectancy either, and it cannot tell anyone what retention does in general. What it can establish is what changed for one person, when, and alongside what else. That is a smaller claim than any study on this page is currently in a position to make, and it is the only one that is about you.