About three months after the first test if the result was abnormal, and as soon as possible if no sperm were found. The UK's NICE guideline says to "undertake repeat confirmatory tests ideally 3 months after the initial analysis to allow time for the cycle of spermatozoa formation to be completed", with the exception that azoospermia or a very low count should be repeated as soon as possible. The American Urological Association and ASRM guideline asks for at least two semen analyses obtained a month apart, because semen parameters "may vary substantially from test to test". In a 2003 study of healthy men sampled weekly, two concentration results from the same man had to differ by about 78% before the change counted as real. So the retest is the next step after an abnormal result, and a single test is never the verdict.
What the guidelines say
NICE's fertility guideline (NG257, 2026) is direct: if the first semen analysis is abnormal, offer a repeat confirmatory test, ideally three months after the first, "to allow time for the cycle of spermatozoa formation to be completed". If a gross deficiency was found, meaning azoospermia (no sperm) or severe oligozoospermia (a very low count), the repeat should happen as soon as possible, because the next steps are different and shouldn't wait. The AUA and ASRM guideline, from the US, puts it as at least two analyses a month apart, especially when the first is abnormal, and explains why: semen measures "are highly variable biological measures and may vary substantially from test to test". The two differ on the gap, not the principle.
Why three months
Sperm take about 74 days to develop from stem cell to finished sperm, on Heller and Clermont's 1963 estimate, and then one to two weeks to pass through the epididymis; a 2006 isotope study put the whole journey at about 64 days on average, with real variation between men. The sample on any day is a mix of cells that started developing across the previous two to three months. A retest three months later measures cells that developed entirely after the first, which is the only way to tell whether that result was a blip or a pattern. The how-long guide goes through the timing.
Why one result is noisy
Alvarez and colleagues (Human Reproduction, 2003) took weekly samples from 20 healthy donor candidates for ten weeks and measured how much each man varied from himself. Concentration varied most: two sequential results from the same man had to differ by 77.8% before the difference could be called real rather than noise. For progressive motility the figure was 49.2%, for morphology 58.0% and for vitality 34.4%. The authors concluded that population reference values have "little diagnostic value because of their marked individuality", while serial results in the same man remain useful. That is the case for the retest in one sentence: your second result is compared with your first, not only with a table.
Fever, illness and the three-month shadow
Carlsen and colleagues (Human Reproduction, 2003) followed 27 healthy men for 16 months with monthly semen analyses and a daily record of fever. A fever during the later stages of sperm development cut sperm concentration by roughly a third, each extra day of fever took off another 7 to 8.5%, and the share of normally shaped sperm fell too, with wide variation between men. The effect appeared in samples taken weeks after the illness, when the men felt fine. If the three months before your first test included a high fever, a bad flu, a course of some medications, a hot-tub habit or a heavy month of drinking, the first result may be describing that month rather than you, and the retest is where you find out.
Keeping the two tests comparable
The retest only tells you something if the conditions match. The WHO manual asks for 2 to 7 days without ejaculation before a sample; a shorter gap lowers volume and count, a longer one lowers motility, so use the same number of days both times and tell the lab what it was. Use the same laboratory where you can, because methods and reporting differ between labs. Collect the whole sample. And write down anything unusual in the three months before each test, so the two results can be read against what was going on. If you have changed a habit since the first test, the retest is also the first sample that can show it; the timeline tool gives you the date.
When not to wait
Do not wait three months if no sperm were found, if the count was very low, or if the result came with a flag (a low volume with a low pH, a high white cell count) or a history that raises one: a varicocele, an undescended testicle as a child, mumps after puberty, groin or testicular surgery, chemotherapy or radiotherapy, or past testosterone or anabolic steroid use. Those are the cases the guidelines send to a urologist or andrologist; the results guide lists them in full.
What to do
- If the first result was below a reference limit, book the retest for about three months later, or ask for two a month apart if your clinic follows the AUA approach.
- If no sperm were found or the count was very low, ask for the repeat straight away and for a referral.
- Match the abstinence window (2 to 7 days, the same number both times), the lab and the time of day.
- Write down fevers, illnesses, medications, heat exposure and heavy weeks in the three months before each test.
- Compare the second result with the first, not only with the table. The Motily app lines each value up across reports, and the checker tool shows each against the WHO limit.
Related guides
Sources
- NICE. Fertility problems: assessment and treatment (NG257). 2026.
- American Urological Association and ASRM. Diagnosis and treatment of infertility in men: AUA/ASRM guideline. 2020, amended 2024.
- World Health Organization. WHO laboratory manual for the examination and processing of human semen, 6th edition. 2021.
- Chung E, et al. Sixth edition of the World Health Organization laboratory manual of semen analysis: updates and essential take away for busy clinicians. Arab Journal of Urology, 2023. (Abstinence of 2 to 7 days.)
- Alvarez C, et al. Biological variation of seminal parameters in healthy subjects. Human Reproduction, 2003.
- Carlsen E, et al. History of febrile illness and variation in semen quality. Human Reproduction, 2003.
- Heller CG, Clermont Y. Spermatogenesis in man: an estimate of its duration. Science, 1963.
- Misell LM, et al. A stable isotope-mass spectrometric method for measuring human spermatogenesis kinetics in vivo. Journal of Urology, 2006.