References and explanations

Every study behind this site, grouped by topic: who was studied, how it was measured, what it found and where we use it. Each topic starts with the terms those studies use, explained in plain English.

How to judge a study

Measured beats self-reported

When a clinician holds the ruler or a couple holds a stopwatch, numbers come out lower and more consistent than when people report on themselves.

Same method, fair comparison

Pressing the ruler to the bone adds over a centimetre compared with measuring at skin level. Only compare numbers taken the same way.

Who took part matters

Random or compulsory samples, like military medical checks, beat volunteers. Big samples beat small ones, but a big biased sample is still biased.

Look past the average

The spread tells you what's normal. A single average hides the fact that most people sit some distance from it.

Size

Go to the Size page

Terms used in size studies

Bone-pressed length: Measured along the top of the erect penis with a rigid ruler pressed into the fat pad until it touches the pubic bone. It's the method the Size page asks you to use, because it cancels out differences in body fat.
Skin-level length: Also called non-bone-pressed. The ruler rests on the skin without pressing into the fat pad, so it reads shorter: by about 1.3 cm on average across clinical studies.
Stretched length: The soft penis gently pulled to its full length and measured. Clinics use it because no erection is needed, but it usually reads shorter than a real erection.
Induced erection: An erection produced in the clinic with an injection, so researchers can measure a full erection accurately.
Girth: The distance around the shaft, usually taken at the middle. The base normally measures a little thicker.
Researcher-measured and self-reported: In researcher-measured studies a health professional does the measuring. In self-reported studies men measure themselves, and the averages come out consistently higher.
Mean and standard deviation (SD): The mean is the average. The SD describes the spread: on a bell curve, about two thirds of men fall within one SD of the mean and 95% within two.
Percentile: The share of men smaller than a given size. At the 90th percentile you're bigger than 90 men out of 100.
Nomogram: A chart of percentiles that doctors use to see where a measurement sits.
Meta-analysis: One figure made by pooling several studies. It's only as good as the studies that go in, and mixing measuring methods can pull the result up or down.
Volunteer bias: People who sign up for a study aren't always typical. Men who are happy with their size may be keener to be measured, which can push averages up.
Micropenis: A medical term for a stretched length more than 2.5 standard deviations below the average for age. In adults that's under about 9 cm stretched, and it's rare.

Every clinic-measured erect study, bone-pressed

Each dot is one study's average, and the line shows where the middle two thirds of its men fell. All were measured by health professionals, pressing to the bone. The dashed lines are for comparison.

Units

Study figures as compiled in calcSD's dataset list. The studies differ in age, country and how men were recruited, which explains much of the spread between them. Every study on this chart is in our average; the two teams with two groups count once each.

How we built the Size page's average

We went through every dataset on calcSD's list and kept a study only if it passed all six checks below. Eleven research teams passed. Their average erect length is 13.9 cm (5.47 in), SD 1.72, and their average girth is 11.5 cm (4.54 in), SD 1.23. Each team counts equally, so one big clinic can't drown out the rest. calcSD weights the same studies partly by size and gets 14.0 cm, so the choice barely matters.

Measured by a professional

A doctor, nurse or researcher held the ruler. Men who measure themselves report bigger numbers.

A real erection

Natural or induced in the clinic. A stretched soft penis doesn't reliably match its erect length, so the two can't be mixed.

Pressed to the bone

For length, the ruler went through the fat pad to the pubic bone, the way the Size page asks you to measure. Skin-level studies count for girth only.

Ordinary adult men

No groups chosen because of worries about size, and none with many under-18s. calcSD had already screened these out.

Full figures published

An average and a standard deviation, so the spread can be pooled properly, not just the middle.

Each team counted once

Where one team published two groups in the same year, we averaged them, in case the same men appear in both.

What we left out, and why

Men measuring themselves

The Kinsey data (1979), Herbenick 2014, Di Mauro 2021 and a 2019 Reddit survey. Self-measurement invites rounding up and attracts men who are proud of their size: the Kinsey data and Di Mauro came out 1.7 and 2.9 cm above our average, and the Reddit survey 4.6 cm above. Herbenick stays on the Size page only as the clearly labelled self-measured option.

Methods never published

LifeStyles Condoms (2001), a condom maker's own measurements of 300 men. The full method was never released, and invented figures attributed to it are still passed around online.

Mixed measuring methods

Veale 2015, the review this site used before, averages bone-pressed and skin-level lengths together, which drags its figure down to 13.1 cm. calcSD also found errors in its erect data. calcSD's corrected rebuild fixes the mixing, but only 191 men in it were measured erect and bone-pressed.

Length measured at the skin

Ali and Ali 2012, Promodu 2007, Yoon 1998 and Wang and Li 1989 are sound studies, but skin-level length reads about 1.3 cm shorter than bone-pressed. We used their girths instead, except Wang and Li, who didn't report one.

Stretched or soft only

38 of the 56 studies in calcSD's main list, including the best-sampled of all: 3,300 Italian 17 to 19 year olds measured at compulsory military checks (Ponchietti 2001). They're good research, but none measured an erection. Some appear on the Myths and records page instead.

The honest caveat

None of the erect studies is a random sample. Most measured volunteers or men attending urology clinics, often with erections induced by injection, and the study averages range from 11.9 to 15.7 cm, more than chance alone would explain. Allowing for that spread, our best estimate is 13.9 cm, give or take about 0.6 cm.

Stamina

Go to the Stamina page

Terms used in stamina and orgasm studies

IELT: Intravaginal ejaculation latency time: the time from penetration to ejaculation. It's what stopwatch studies measure, and it doesn't include foreplay.
Stopwatch study: Couples time themselves at home over several weeks, usually with the partner holding the stopwatch.
Hidden timer: A timer the couple can't see during sex, used to check that watching the clock doesn't change the result.
Median: The middle value when everyone is lined up. Timings are lopsided, with a few very long ones, so the median describes a typical man better than the mean.
Geometric mean: An average suited to lopsided data that reduces the pull of very large values.
Lifelong premature ejaculation: Ejaculating within about a minute of penetration nearly every time since first having sex, with little control and real distress.
Acquired premature ejaculation: A bothersome drop to about three minutes or less, after previously lasting longer.
Delayed ejaculation: Finding it very difficult or impossible to ejaculate despite wanting to.
Refractory period: The recovery time after ejaculation before another erection and orgasm are possible.
Orgasm gap: The difference between how often men and women orgasm during sex. It's widest between straight men and straight women.
Placebo-controlled trial: A trial that compares a treatment with a dummy version, so the difference shows what the treatment really does.
Umbrella review: A review of reviews that pulls together the strongest summaries of the evidence.
Off-label: A medicine prescribed for something other than what it's licensed for. It's legal and common when the evidence supports it.

Sex

Go to the Sex page

Terms used in frequency surveys

Nationally representative survey: A survey designed so the people asked mirror the whole population, usually through random sampling.
General Social Survey (GSS): A long-running survey of US adults, carried out since 1972.
Natsal: Britain's National Survey of Sexual Attitudes and Lifestyles, carried out around 1990, 2000 and 2010.
Sexual frequency: How often people report having sex in a set period, usually the past month or year. Surveys mostly leave it to people to decide what counts as sex.
Sexual inactivity: No sex at all in the past year.
Cohort effect: A difference linked to when people were born, rather than their current age or the year of the survey.
Correlation: Two things that rise and fall together. More sex and more happiness go together, but that alone doesn't show which causes which.

Pubic hair

Go to the Pubic hair page

Terms used in pubic hair studies

Grooming: Trimming or removing any pubic hair, by any method. Some studies count trimming, others only removal, which is one reason their figures differ.
Lifetime and past month: Whether someone has ever groomed, against whether they did so in the last 30 days. The first is always the bigger number.
Nationally representative: Participants are picked at random, for example from postal addresses, and weighted so the results reflect the whole adult population rather than volunteers.
Odds ratio: How much more likely something is in one group than another. An odds ratio of 2 means double the odds; 1 means no difference.
Linked, not caused: A survey can show that grooming and infections go together, but not that one causes the other. People who groom may differ in other ways.
Skin-to-skin infections: Infections spread by skin contact, such as herpes, HPV and syphilis, as opposed to those spread through body fluids.

Myths about size

Go to the Myths about size page

Terms used on the Myths about size page

Standard deviations from average: How far a value sits from the average, counted in SDs (also called a z-score). Beyond about four, a bell curve predicts almost no one.
Extrapolation: Using a bell curve to estimate rare extremes, like 1 man in a million. Real extremes are often more common than the curve predicts, so these are rough guides.
Verified: On this site, verified means measured by someone independent and published where others can check it.
Peer review: Checks by independent experts before a study appears in a journal. It catches many problems, but not all.
Systematic review: A study that gathers all the research on a question using set rules, rather than picking favourites.
Small study: A handful of participants can show that something happens, but not how common it is. Treat findings from very small studies as provisional.

No studies match that search. Try a country, an author or a word like "stopwatch".

The home page quiz

The crowd results on the home page come from visitors' guesses, saved anonymously when they press Enter. Only the three numbers and the date are kept. They show what people expect, not what's true, and they aren't research: anyone can take part, and nobody is checked.

Where this list comes from

Studies are chosen by the authors of this site from peer-reviewed journals, official guidance and national surveys. Several size studies, marked "Added from calcSD", came from the dataset list at calcSD (calcsd.info), an independent project that compiles published penis measurements and separates them by measuring method. Figures marked "via calcSD" are as compiled there. Which studies feed the Size page was decided by our own checks, set out in the Size section above. Links go to the study itself where one is available, or to a Google Scholar search for it.