Male infertility tests: what the WHO 2021 semen figures do and do not mean

The numbers at the bottom of a semen report are fifth centiles, not a pass mark. They describe where the lowest 5% of men whose partners conceived naturally within a year happened to fall (WHO laboratory manual, 6th edn, 2021, Table 8.3), and WHO says in section 8.1.3 of that manual that this fifth percentile "does not represent a limit between fertile and infertile men". A result below one of them does not make a man infertile, and results above all of them do not make him fertile.

Written by Dr. A. Ameer Jahan, Senior Consultant · Reproductive Medicine, Male Infertility & STD, Chairman, AJSMC, Egmore, Chennai (TNMC 28017). This page is about what the tests in a male fertility workup measure, and what each can and cannot tell you. The measurements follow the WHO laboratory manual for the examination and processing of human semen, sixth edition (World Health Organization, 2021). Everything beyond the semen analysis follows the AUA/ASRM Male Infertility Guideline (2020, amended 2024), NICE guideline NG257 (2026) and the EAU Guidelines on Sexual and Reproductive Health, male infertility chapter, March 2026 update, with the Urological Society of India guideline (Indian Journal of Urology, 2023) where an India-specific point exists.

Some of what brings men to a page like this is not an appointment at all. Sudden severe pain in a testicle, with or without vomiting, may be torsion, and the window for saving the testicle is measured in hours. An erection lasting more than four hours, which is priapism, and an inability to pass urine both need care within hours rather than at the next appointment. A hot, swollen, painful scrotum with fever needs care the same day. Any of these means 108, or the nearest hospital with a 24-hour emergency department and an operating theatre. AJSMC is a multi-speciality centre with no casualty unit, no emergency department and no critical care: its ten beds take planned, stable admissions, with a nurse on site overnight and no doctor on the premises. Consultations run Monday to Saturday, 10am to 9pm, and it is closed on Sundays. Almost nothing in a fertility workup is an emergency — torsion at midnight is — though a few findings do need to be seen sooner than the next free slot, and those are listed at the end of this page.

What do the numbers on a semen report actually mean?

They come from one table. WHO's sixth edition pooled semen results from men in twelve countries whose partners had a confirmed natural conception within twelve months, and reported the fifth centile of each measurement.

Source: WHO laboratory manual, sixth edition, 2021, Table 8.3, page 213. AUA/ASRM (2024) reproduces all seven in its Table 4 and NICE NG257 (2026) in recommendation 1.17.1.

The population is men whose partners conceived naturally within a year, so these are figures taken from fertile men rather than a line drawn between two groups. WHO deliberately stopped calling them reference limits: its summary of changes says the distribution of results from presumed fertile men "is not sufficient to establish clinically useful decision limits", and section 8.1.2.2 says such limits still need to be developed. WHO has published none. The sample size also differs by row. Vitality rests on roughly a third as many men as the others and carries the widest confidence interval of the seven, so it does not deserve equal billing with the rest. NICE adds an eighth value, a semen pH of 7.2 or more, which is a carried-forward consensus figure rather than a fifth centile and so sits outside the table.

The reference population has been criticised, fairly. A reappraisal in Human Reproduction (Paffoni and colleagues, 2022) found median sperm concentration varying between 11 and 36 million/ml across the studies WHO pooled, and concluded the aggregated data cannot be treated as coming from one population. No Indian dataset contributed. One further confusion is local: the sixth edition dropped labels such as oligoasthenozoospermia as classification, but many Indian laboratories still print them. Such a word describes one sample on one day, not the man.

How do the 2021 figures differ from the 2010 ones?

They moved in both directions, which makes an old chart misleading rather than merely dated.

Sources: Cooper TG and colleagues, Human Reproduction Update, 2010, for the fifth edition; WHO sixth edition, 2021, Table 8.3, for the current figures. A concentration of 15.5 million/ml sits above the 2010 figure and below the 2021 one. A progressive motility of 31% sits below the 2010 figure and above the 2021 one. Same sample, opposite readings, depending only on which edition it was compared against. If you are holding a report from a few years ago, check which edition it names before drawing anything from it.

How is the sample supposed to be collected?

The collection rules are part of the test, and a sample collected badly produces a number that looks real and means nothing. WHO specifies a minimum of two days and a maximum of seven days of abstinence, and is candid that this rests on clinical experience rather than trial evidence, and that the width of the window itself adds to the variability of the result. The Urological Society of India says no clear recommendation on the ideal abstinence period can be made from the conflicting evidence, before offering two to five days as practical.

The rest, from section 2.3.1 and the collection instructions in 2.3.2: the sample should ideally be produced in a private room close to the laboratory; examination should begin within 30 minutes and in any case within 60; a sample collected elsewhere must be kept between 20 °C and 37 °C in transit, carried close to the body under clothing, and delivered preferably within 30 minutes and, in the manual's own words, "at least no longer than 50 minutes after collection"; the whole ejaculate must be collected and any lost fraction recorded, because volume and total sperm number are both wrong without it; ordinary latex condoms must not be used and lubricants should be avoided.

Vitality is not run on every sample. WHO's template triggers it when fewer than 40% of sperm are motile, and requires it within an hour of ejaculation. It answers one question: are the immotile sperm dead, or alive but not swimming? Low motility with preserved vitality points towards a structural problem with the tail, a different line of investigation. On home and postal kits, AUA/ASRM notes accuracy varies and there is no substitute for a specialised andrology laboratory.

Why is one abnormal result usually repeated, and how soon?

Because semen quality varies substantially between samples from the same man. AUA/ASRM puts it directly: these are highly variable biological measures, and statement 9 (Expert Opinion) adds that results carry most weight when several abnormalities appear together rather than one in isolation. WHO is more careful than the common instruction always to repeat. Section 2.2 says a single examination is enough to decide the next steps, while noting that defining an exact baseline for one man may need two or three samples. The bodies then disagree on the interval, and the disagreement is worth seeing rather than settling with one confident number.

The half of NICE's recommendation usually dropped from summaries is the exception inside it: where no sperm at all are found, the three-month wait does not apply. On when a couple should be assessed at all, NICE advises investigating both partners after a year of unprotected intercourse without conception, and referral at presentation where the female partner is 36 or over or either partner has a known or suspected cause. Section 2(j) of the Assisted Reproductive Technology (Regulation) Act, 2021 uses the same one-year period. The Urological Society of India names the specific Indian failure plainly: the male partner is bypassed in the majority of infertile couples.

When are blood hormone tests added?

Not routinely, and not at the first appointment for most men. NICE recommendation 1.17.5 gives the cleanest trigger: for men with two or more abnormal semen analyses, offer physical examination of the scrotum and testes, and consider measuring serum testosterone and gonadotrophin levels.

The panels differ. AUA/ASRM statement 10 (Expert Opinion) names FSH and testosterone only, and among the findings that trigger it are oligozoospermia or azoospermia, small testes and signs of a hormonal problem on examination. LH and prolactin are not in that first-line panel. The EAU is broader, directing total testosterone with FSH and LH in all cases of oligozoospermia and azoospermia (Strong). The Urological Society of India is narrower, offering an endocrine profile only where the semen analysis is abnormal on at least two occasions (Expert Opinion, Strong). Examination is not optional either, and in India it carries an extra purpose: the Urological Society of India specifies that it should help rule out infections including genito-urinary tuberculosis, a real cause of obstruction here that does not appear in Western guidance.

What does it mean when no sperm are found at all?

Azoospermia is a laboratory finding, not a diagnosis. WHO states it plainly: it remains a description of findings in the ejaculate and not a diagnosis or a basis for therapy, and the term can only be used where no sperm are found in the sediment of a centrifuged sample. The manual attaches three cautions. Whether sperm appear in that pellet depends on how long and how fast the sample was spun and how much of the pellet was examined; spinning at 3000 g for 15 minutes does not pellet every sperm; and centrifugation can destroy motility and underestimate concentration, so a centrifuged count must be labelled as such.

The next question is whether sperm are not being made, or are being made and not getting out, and that is settled clinically first. AUA/ASRM statement 11 (Expert Opinion) evaluates azoospermia initially with physical examination, semen volume, semen pH and serum FSH; statement 21 (Expert Opinion) advises against routinely performing a diagnostic testicular biopsy to tell the two apart.

The figures behind that advice need their conditions attached. In azoospermic men, AUA/ASRM states that an FSH above 7.6 IU/L with a testis longitudinal axis under 4.6 cm indicates an 89% likelihood that the cause is impaired sperm production, and an FSH below 7.6 IU/L with a testis axis above 4.6 cm a 96% likelihood of obstruction. That is graded Expert Opinion, the weakest tier, from one retrospective series of 153 azoospermic men (Schoor and colleagues, Journal of Urology, 2002). The original study joined its two criteria with "or" rather than "and", and reported them the other way round: those were proportions of men already known to have each diagnosis, not the chance of a diagnosis given a result. The EAU declines to use these cut-offs at all, looking instead for a testis volume above 15 ml with a normal FSH, and warning that men with maturation arrest can have both. The usable message is that a biopsy is rarely needed to answer the question, not that these numbers answer it.

Distal obstruction is suspected where the ejaculate volume is low with acidic semen below pH 7.0, and AUA/ASRM notes that a man with a normal semen pH is unlikely to have a complete one. Seminal fructose, still bundled into some panels, is described by AUA/ASRM as relatively unreliable and not necessary, and the Urological Society of India says independently that it is not a definitive marker for separating the two causes.

Which genetic tests are done, and at what sperm concentration?

Two different tests with two different thresholds. Collapsing them into a single "five million" figure is the commonest error on this topic, and an understandable one, because the American threshold for the second test was five million until the 2024 amendment moved it.

AUA/ASRM applies both concentration thresholds only alongside raised FSH, testicular atrophy or a diagnosis of impaired sperm production. NICE restricts both to cases that are otherwise unexplained. The EAU keeps a discretionary band between one and five million for microdeletion testing, stating that no absolute threshold can be given universally. The change came from a meta-analysis finding microdeletions in 5% of men with concentrations of 0 to 1 million/ml against 0.8% between 1 and 5 million/ml. India has no equivalent national threshold document; this guidance is American, European and British.

The microdeletion test comes before any surgical planning for a specific reason. Where AZFa or AZFb is completely deleted, sperm have not been retrieved by microdissection testicular sperm extraction. The EAU advises against testicular sperm extraction in that situation (Strong), and NICE recommendation 1.26.1 is flat: do not offer surgical sperm retrieval in the presence of a Y chromosome AZFa or AZFb microdeletion. A result of that weight is also why azoospermia should never be handed over on paper or read out over the phone without a consultation booked around it. Where the vas cannot be felt, AUA/ASRM statement 25 (2020, amended 2024) adds renal ultrasonography, because between 26% and 75% of men with unilateral absence of the vas have a kidney anomaly on the same side.

What does the evidence say about repairing a varicocele?

Varicoceles are common. AUA/ASRM records them in roughly 15% of adult men and 40% of infertile men, about 85% on the left. The honest position on repair separates what is well evidenced from what is not, and the certainty grades do that better than any summary sentence.

Source: Persad E and colleagues, Cochrane Database of Systematic Reviews, 2021, Issue 4, CD000479.pub6, searches closing 4 April 2020. The review as a whole covered 48 randomised trials and 5,384 participants, but those totals are spread across four comparisons and are not the evidence base for any single row above. The live-birth row is the one to look at hardest: two trials, 204 participants between them, pointing in opposite directions, with an interval running from 0.19 to 26.93. The reviewers conclude that it remains uncertain whether any treatment compared with no treatment is of benefit on live birth. Where the review is most confident is about which operation, not about whether to operate, and for that technique comparison there is no live-birth evidence at all.

AUA/ASRM statement 26 advises considering surgical varicocelectomy in men attempting to conceive who have a palpable varicocele, infertility and abnormal semen parameters, azoospermic men excepted. That is a Moderate Recommendation at Grade B, resting substantially on non-randomised studies, a limitation the guideline states about its own evidence. The EAU cites the same Cochrane figures and still recommends treatment strongly, where the varicocele is clinical, the semen parameters abnormal, the infertility otherwise unexplained and the female partner has good ovarian reserve. NICE recommendation 1.28.1 uses the weakest verb of the three, consider. India has no equivalent graded national guidance.

All three agree on where repair does not belong. AUA/ASRM statement 27 (Strong Recommendation, Grade C) advises against varicocelectomy for varicoceles found only on imaging and not felt on examination, citing a systematic review of 7 trials in 548 participants that found no demonstrable benefit in pregnancy or bulk semen parameters, beyond a possible small effect on progressive motility that the reviewers judged unlikely to matter clinically. The EAU says the same about a subclinical varicocele and about any man whose semen analysis is normal (Strong). In non-obstructive azoospermia, AUA/ASRM statement 28 (Expert Opinion) requires that couples be told of the absence of definitive evidence supporting repair before surgical sperm retrieval, noting that up to 35% of such men have sperm detected on a later semen analysis with no intervention at all, and that repair defers assisted reproduction by at least six months. Varicocele surgery is not performed at AJSMC; where it is being considered that is a referral onward, and this conversation belongs before the referral, not after.

Do smoking, alcohol and weight really affect sperm count?

The list is shorter than the internet suggests, and for several popular culprits the honest answer is that nobody has assessed them properly rather than that they are harmless.

AUA/ASRM statement 42 is a Clinical Principle that clinicians should not prescribe exogenous testosterone therapy to a man interested in current or future fertility, and NICE recommendation 1.24.3 tells prescribers not to offer androgens to treat semen abnormalities.

On supplements the evidence does not support the confidence with which they are sold. AUA/ASRM statement 45 (Moderate Recommendation, Grade B) says their benefits are of questionable clinical utility and existing data are inadequate to recommend any specific agent. The Cochrane review of antioxidants (de Ligny and colleagues, 2022, CD007411.pub5) covered 90 randomised trials and 10,303 subfertile men, examined around twenty agents and could recommend none. Its live-birth analysis rests on 12 trials in 1,283 men at very low certainty and its clinical-pregnancy analysis at low certainty, and when trials at high risk of bias were excluded the evidence of benefit disappeared. NICE recommendation 1.24.6 goes furthest: do not offer supplements, antioxidants or medical treatments to improve sperm DNA integrity.

Which tests do you not need at the first appointment?

Several of these are bundled into fertility packages sold across India, and none of them belongs in a first workup.

When should you see a doctor?

Book an outpatient appointment if you and your partner have been trying for a year without conceiving, or sooner if your partner is 36 or over, if either of you has a known cause or a history that predisposes to one, if there have been two or more pregnancy losses, or if an assisted reproduction cycle has failed. Come together if you can; the assessment of both partners runs in parallel, and a workup starting with only one of you takes longer to reach an answer.

Bring every previous semen report with its date, and check whether each names the WHO edition it was compared against. Bring hormone results with their dates, and any regular medication, which the doctor will ask about by name, including anything bought without a prescription. NICE recommendation 1.13.1 asks the doctor to take that medication history. The question is not a moral one, and the answer changes what the tests mean.

Say so when you call, and ask to be seen sooner, if you have felt a firm painless lump in a testicle, if a swelling in the scrotum is new, large, cannot be reduced when you lie down or is on the right side alone, if you have breast enlargement or nipple discharge along with headaches or any change in your vision, or if a report has come back with no sperm on it. Each of those is a different pathway from a routine fertility appointment.

Do not come here, or wait for opening hours, for sudden severe testicular pain, an erection lasting more than four hours, an inability to pass urine, or a hot swollen scrotum with fever. Call 108 or go to the nearest hospital with a 24-hour emergency department. AJSMC consultations run Monday to Saturday, 10am to 9pm, and it has no emergency department and no critical care; its ten beds take planned, stable admissions only. Surgical sperm retrieval, varicocele surgery and assisted reproduction are not performed here, and would be a referral onward to a centre that provides them.

An infertility diagnosis carries real psychological risk, and in the social setting most couples in Chennai are managing, the pressure can be severe. If you or your partner are feeling hopeless, or having thoughts of harming yourself, that needs help the same day from a service with out-of-hours cover, not an appointment slot at a centre that closes at 9pm. Call 112, or 108 for an ambulance in Tamil Nadu, or go to the nearest hospital with a 24-hour emergency department. Tele-MANAS on 14416 is free, government-run and answers 24 hours a day, every day. For the fertility assessment itself, and only for that, the outpatient at AJSMC on Police Commissioner Office Road, Egmore, Chennai 600008 can be reached on 044 2532 2021, 9150615999 or 9150642999.