Semen analysis: how to prepare and how to read the result (WHO 2021)

The most important part of preparation is the period of sexual abstinence: according to the WHO recommendation it should be 2 to 7 days, and the exact number of days should be given to the laboratory. The sample is usually collected by masturbation into a sterile container supplied by the laboratory, in a room at the facility or, if the laboratory allows, at home with prompt delivery. It is important to collect the whole ejaculate, including the first portion, which contains the most sperm; if part of the sample was lost, the staff should be told, because the result must then be interpreted with caution. Ordinary condoms and lubricants, which can be toxic to sperm, must not be used.

The sample should be delivered to the laboratory soon after collection (usually within about an hour; the laboratory gives the exact rules) and kept at a temperature close to body temperature, for example in an inner pocket, not in a cold car or a refrigerator. Sperm motility falls quickly with time and falling temperature, so a delay distorts the result. Before the test it is worth telling the laboratory about any recent fever or illness, medicines and supplements being taken, procedures and exposure to high temperature (sauna, work in the heat), because all of these can affect the result.

A second, more recent meta-analysis (Raditya et al., 2025; 22 studies, 31,640 samples) recalls that WHO recommends abstinence of 2–7 days and tested whether shorter abstinence (under 2 days) improves parameters. Short abstinence reduced volume (mean difference −0.83), sperm concentration (−8.39) and DNA fragmentation (−3.82), and increased the proportion of sperm with total motility (+2.30) and progressive motility (+2.18); morphology did not differ. The authors suggest that in assisted reproduction techniques shorter abstinence may be worth considering, but for a standard diagnostic semen analysis the 2–7 day range applies.

The practical conclusion for the person being tested is simple: a difference of a few days of abstinence can change concentration and volume even without any change in spermatogenesis itself. For comparability of successive results one should therefore keep a similar interval, rather than 'improve' the result with a very long or very short abstinence. Both meta-analyses rest on studies of varied quality and in different populations (some involved men with fertility problems), so their results concern trends, not a single sample.

The values in the sixth edition of the WHO manual (2021) come from an analysis of data from men whose partners became pregnant within 12 months (so-called time to pregnancy ≤12 months). They cover 3,589 men: the data from the fifth edition (1,800 people) supplemented by new studies from 2010–2020 (1,789 people). The numbers given are the 5th centile: the value below which 5% of men in this population lie. This means that even among fertile men about 1 in 20 will have a result below this limit for a given parameter.

The most important sentence in the literature is that reference values are not a diagnostic threshold. The critical review of the sixth edition (Boitrelle et al., 2021) points out that the new manual has abandoned the old descriptive terms such as normozoospermia, asthenozoospermia or teratozoospermia, and that the 5th centile values are treated only as one way of interpreting results, insufficient to diagnose infertility. Björndahl (2022) states plainly that the main problem was that these ranges were misread as distinct limits between fertility and infertility.

Why does this matter? Because the distributions of parameters in fertile and subfertile men overlap strongly. A result of 14 million/ml does not mean infertility, and 25 million/ml does not guarantee conception, because the chance of pregnancy also depends on the partner's fertility, her age, the frequency and timing of intercourse, and the quality of sperm, which a standard semen analysis does not fully measure (for example DNA integrity). Clinicians therefore assess not one number but the whole picture: several parameters at once, repeated tests, the history of trying and information about the partner.

A practical rule follows from both studies: between two successive results from the same man, sperm concentration may differ by tens of percent without any change in his health. According to the study of healthy donor candidates, the difference between successive concentration measurements had to exceed about 78% to be considered significant, so a fall from 40 to 28 million/ml (about 30%) lies within natural noise and is not proof of deterioration; in men with fertility problems, variation in Keel's study was at least as large. A single result that departs from the norm is particularly unreliable: the greater the variability, the greater the chance that the next test will come out better (or worse) purely by chance. For this reason doctors usually recommend repeating an abnormal test, often after a few weeks or about three months, because the spermatogenesis cycle lasts several dozen days and an earlier illness, fever or procedure may have lowered the result temporarily.

Morphology deserves particular attention. The 4% threshold for normal forms means that in the semen of fertile men the large majority of sperm look 'abnormal' by strict criteria, which is normal. Assessment of morphology is also the most subjective part of the test and depends greatly on the laboratory's experience and staining method: a low value in one laboratory need not be the same in another. Sperm concentration and motility, in turn, depend on many external factors, so interpreting them without taking abstinence and delivery time into account can be misleading.

A complete absence of sperm in the ejaculate (azoospermia) needs confirmation: the laboratory should examine the pellet after centrifugation, and the test is usually repeated. The diagnosis matters clinically because it distinguishes obstruction (sperm are produced but do not reach the ejaculate) from impaired production, which needs different work-up, including hormone measurement (FSH, LH, testosterone) and a scrotal ultrasound. This goes beyond the scope of this article and is a matter for an andrologist.

A result departing from the reference values is not a diagnosis, only a signal for further steps. The first is to repeat the test under similar conditions (abstinence, transport of the sample), preferably in the same laboratory to limit methodological differences. If the abnormality persists, the andrological history covers the time of trying for a child, earlier pregnancies, past illnesses (mumps, infections, testicular injuries, hernia operations), exposures (high temperature, solvents, smoking, alcohol), medicines — especially testosterone and anabolic steroids, which suppress spermatogenesis — as well as body weight and habits.

Typical additional tests are hormone measurements (FSH, LH, testosterone; we cover their interpretation in the entry on sex hormones and in a separate article on high FSH in men), a scrotal ultrasound assessing the testicles and the veins of the spermatic cord, and in selected cases genetic tests (with a very low sperm count or none) and semen cultures when infection is suspected. It is worth remembering that assessing the couple's fertility also requires assessing the partner: many couples with a semen result slightly below the norm conceive naturally, and a couple with results within the norm may have problems on the woman's side. The standard moment to start investigation is regarded as no pregnancy after 12 months of regular trying, and earlier when risk factors are known or the partner is older.

A semen analysis is a very useful test, but only if it is performed and read with due care. Before the test it is worth ensuring 2–7 days of abstinence and proper collection and prompt delivery of the sample. After the test, read the result as a set of numbers together with their variability, not as a verdict. Two mistakes are the most harmful: being reassured by a result 'within the norm' despite long attempts to conceive, and being alarmed by a single result below the norm that months later turns out to be normal. The evidence we rely on in this article comes from observational studies, meta-analyses and narrative reviews, and the WHO source data come from a population of fertile men.