How long it lasts: the clock, the worry and what helps

The stopwatch says 5.4 minutes, but only for straight couples. Here is what is known about gay men, where too fast really begins, and which tricks and treatments have evidence behind them

Porn has done men no favours here. Scenes are edited, shot over hours and finished on cue, and somewhere along the way "lasting" turned into a performance target. The real numbers are smaller, more reassuring and, for gay men, frustratingly incomplete.

This chapter is about the clock: how long men actually last, where "too fast" really begins, how many men are there versus how many worry about it, and what has been shown to buy you extra minutes.

The stopwatch study

In the early 2000s a Dutch psychiatrist, Marcel Waldinger, did something nobody had done properly: he handed out stopwatches. Five hundred couples from the Netherlands, the UK, Spain, Turkey and the United States timed their sex for four weeks, from penetration to ejaculation. They were ordinary couples, not patients.

The median was 5.4 minutes. The spread was huge, from about half a minute to 44 minutes, and the curve was lopsided: most men bunched up in the first few minutes and a long thin tail stretched off to the right. The median dropped with age, from 6.5 minutes in men aged 18 to 30 to 4.3 minutes over 51, and Turkey came in lowest at 3.7. Condoms made no difference to the median, and neither did circumcision outside Turkey.

And between men?

The data on gay men are surveys, not stopwatches. A 2019 meta-analysis pooled the four studies that compared gay and straight men directly: gay men had about 28% lower odds of reporting premature ejaculation, but the studies were few, used different questionnaires and drew on volunteers.

The largest single comparison, an online survey of 3,878 men in 2022, found probable or definite premature ejaculation in 16% of gay men against 20% of straight men. Once other factors were taken into account, the gap disappeared. More interesting: gay men with the problem reported longer typical latencies and less distress than straight men with it, and longer latencies went with anal sex itself: men who topped lasted longer than men having vaginal sex. The American urologists' guideline sums it up: measured properly, the prevalence in men who have sex with men looks about the same as in straight men.

Why topping might take longer is not known. A different kind of grip and friction is a reasonable guess, but nobody has tested it. For masturbation and oral sex, we found no reliable timings at all.

When "fast" becomes a problem

The International Society for Sexual Medicine definition has three parts, and all three are needed:

• Time. Coming always or nearly always within about 1 minute of penetration since your first sexual experiences (lifelong), or a clear, bothersome drop to about 3 minutes or less after years of lasting longer (acquired).
• Control. Being unable to hold back on all or nearly all occasions.
• Consequences. Distress, frustration, or avoiding sex because of it.

The definition literally says "vaginal penetration", because that is where the evidence came from. The 2022 survey concluded that the same three criteria work for gay men, with some allowance for their longer times and lower distress.

Now compare the worry with the condition. In an online survey of more than 12,000 men in the US, Germany and Italy, 22.7% said they had poor control over coming and were bothered by it. A Turkish field survey of 2,593 couples found the same thing from another angle: 20% of men complained of coming too fast, but only 2.3% fitted the lifelong pattern and 3.9% the acquired one. The biggest groups were men who came fast only now and then (8.5%) and men who felt quick while their timing was normal (5.1%). Doctors class the second group as subjective premature ejaculation, and its description explicitly includes men who last more than 5 minutes.

What actually works

Stop-start and squeeze. The stop-start method was described by a urologist in 1956: get close, stop, let the urge fade, start again. Masters and Johnson added the squeeze: firm pressure just below the head until the urge passes. They are still the most used behavioural treatments. Masters and Johnson claimed a 97.8% success rate that nobody has managed to replicate, and a Cochrane review called the evidence for psychological treatments weak and inconsistent. Other studies report success in 60% to 90% of men, and the American guideline advises that combining behavioural methods with medication may work better than either alone. The training is also good practice for the point of no return: learning where it is, and stopping just before.

Pelvic floor training. In an Italian study, 40 men with lifelong premature ejaculation spent 12 weeks on pelvic floor training, three hours a week: clenching exercises, biofeedback and mild electrical stimulation through an anal probe. Average time went from well under a minute to about 2.5 minutes, and 33 of the 40 gained control. There was no comparison group, so take it as promising rather than proven.

Numbing sprays and creams. A lidocaine-prilocaine spray applied to the head five minutes before sex raised average time in a three-month trial from about half a minute to 2.6 minutes, against 0.8 on placebo. A meta-analysis found topical anaesthetics more effective than placebo, though most trials were of uncertain quality. Two catches. Too much numbs the wrong thing: penile numbness and softer erections are listed side effects. And it rubs off: the guidelines warn of numbness in the partner unless a condom is used. Those warnings were written about vaginas, but the lining of the rectum absorbs drugs well, which is why suppositories work, so a bottom who suddenly feels less has a likely culprit. A condom over the treated cock solves it.

Condoms. In Waldinger's study ordinary condoms did not change the median time, so do not expect a regular rubber to do the job. Condoms with benzocaine inside are a different product: in a 2025 Turkish trial they lengthened time, less than spray or cream but with the fewest side effects.

Dapoxetine. A short-acting SSRI taken one to three hours before sex. In two large trials, men starting at under a minute reached 2.8 minutes on 30 mg and 3.3 minutes on 60 mg, against 1.75 on placebo, with nausea in one man in five at the higher dose. It is approved in some countries and not in others, and in one clinic 90% of men had refused it or stopped within a year.

Daily SSRIs. Antidepressants such as paroxetine, sertraline and fluoxetine, taken every day, delay ejaculation more than on-demand pills. Paroxetine was the strongest in a meta-analysis, with an increase of about 8.8 times. The price: possible tiredness, lower desire and erection trouble. They need a prescription and a conversation with the doctor about those trade-offs.

The second round. Wanking beforehand and simply having sex more often are among the tricks men with the problem use most, along with changing position and pausing. No trial has tested coming once and going again, but plenty of men swear the second round lasts longer. How long you wait in between is the subject of Going again.

The opposite problem

Some men take too long, or cannot come at all with a partner. Delayed ejaculation affects about 1% of men lifelong and about 5% who develop it later, and becomes more common with age. The usual suspects are medications, especially SSRIs (the same drugs that help men who come too fast), plus some blood pressure and prostate drugs, and alcohol. Habits matter too: frequent, vigorous or idiosyncratic masturbation is a recognised factor, the death grip discussed in Sensitivity.

Many men who cannot come with a partner have no trouble alone, which points at the head rather than the plumbing. If it started with a new medicine, the guideline answer is to change the dose or the drug rather than suffer. And if it only happens with one new partner, give it a few nights before deciding anything is wrong.