ICSI explained: when injecting a single sperm helps, and when it adds nothing

What ICSI is

Intracytoplasmic sperm injection (ICSI) is a laboratory step inside an IVF cycle. In conventional IVF, each collected egg is placed in a dish with tens of thousands of prepared sperm, and one of them has to bind to the egg's outer shell and get through it unaided. In ICSI, an embryologist holds the egg under a microscope, picks up a single sperm in a very fine glass needle and injects it straight into the cytoplasm, the inside of the egg.

Everything else is the same as in IVF: ovarian stimulation, egg collection, culture of the embryos for a few days, transfer of one embryo and freezing of any others. The Human Fertilisation and Embryology Authority (HFEA) puts it simply: the treatment is exactly the same as IVF, the only difference being that a single sperm is injected into the egg instead of the eggs and sperm being left to fertilise (HFEA, Intracytoplasmic sperm injection (ICSI)). ICSI is therefore not an alternative to IVF but a way of carrying out the fertilisation step.

Where it came from

The technique was developed at the Centre for Reproductive Medicine of the Vrije Universiteit Brussel. In July 1992, Gianpiero Palermo, Hubert Joris, Paul Devroey and André Van Steirteghem reported in The Lancet the first pregnancies after injecting a single sperm into an egg, in couples with severely impaired sperm in whom IVF and an earlier micro-injection method had failed (Palermo et al., Lancet 1992;340:17-18, doi:10.1016/0140-6736(92)92425-F). Of 47 injected eggs, 31 were fertilised; four pregnancies followed, and four healthy babies were born.

For men with very low sperm counts, poorly moving sperm, or sperm that could only be obtained surgically, this changed what was possible: one viable sperm per egg is enough. The broader history of assisted reproduction since the first IVF birth is told in IVF from 1978 to today.

Who it is recommended for

The HFEA lists the situations in which a doctor may recommend ICSI:

• a very low sperm count, or sperm that are abnormally shaped or do not move normally;
• a previous IVF cycle in which none, or very few, of the eggs were fertilised;
• sperm that have to be collected surgically from the testicles or the epididymis, for example after a vasectomy or because of a blockage;
• frozen sperm of limited quality, particularly sperm stored because fertility was under threat;
• embryo testing for a genetic condition, where extra sperm stuck to the outside of the egg could interfere with the results.

It also says when ICSI is unlikely to help: if fertilisation failed in an earlier cycle because the eggs were of poor quality or immature, injecting the sperm does not solve that problem. Because sperm counts vary from one sample to the next, clinics usually repeat an abnormal semen analysis before drawing conclusions; our article on whether sperm counts are falling explains what the World Health Organization reference values mean.

Most IVF is now ICSI

Use of the technique has spread far beyond the problem it was designed for. The European Society of Human Reproduction and Embryology (ESHRE) states in its 2025 fact sheet on assisted reproduction that ICSI is the most common fertilisation technique and accounts for around 73% of all treatments worldwide, with large differences between countries, even though outcome rates with the two techniques are comparable. In ESHRE's European data for 2020, the mean pregnancy rate per embryo transfer was 33.2% after IVF and 33.0% after ICSI.

Male infertility alone does not explain this. Clinics and patients have often chosen ICSI as insurance against total fertilisation failure, the disappointing case in which none of the eggs fertilise in conventional IVF, or simply because it has become routine laboratory practice.

What the trials show without male infertility

The question is whether ICSI improves the chance of a baby when the sperm are normal. The largest randomised trial on the subject was published in The Lancet in April 2021 by Vinh Q. Dang and colleagues, working at two IVF centres in Ho Chi Minh City (doi:10.1016/S0140-6736(21)00535-3). They randomly assigned 1,064 couples in which the male partner had a normal sperm count and motility to ICSI or conventional IVF.

• Live birth after the first embryo transfer: 184 of 532 couples (35%) in the ICSI group and 166 of 532 (31%) in the IVF group. The difference was not statistically significant (risk ratio 1.11, 95% confidence interval 0.93 to 1.32).
• Total fertilisation failure: 29 couples (5%) with ICSI and 34 (6%) with IVF, again with no significant difference.

The authors concluded that, for these couples, ICSI did not improve the live birth rate and that their results challenge its routine use. The trial was open-label (couples and embryologists knew which group they were in) and was run in one country, so it does not settle every question, but it addressed the outcome that matters most, a live birth, rather than fertilisation rates alone. Why that distinction matters is explained in IVF success rates explained.

The HFEA reaches the same position: there is currently no scientific evidence to support ICSI for infertility that is not related to sperm, and professional guidance from the British Fertility Society and ESHRE does not recommend it over standard IVF in these circumstances. It adds that choosing ICSI when it is not needed may increase the cost of treatment without clear evidence of a better chance of success.

What NICE's 2026 guideline says

NICE's guideline Fertility problems: assessment and treatment (NG257), published on 31 March 2026, replaced a 2004 recommendation on ICSI that its committee described as outdated. Section 1.50 now reads as follows (NICE NG257, section 1.50):

• Offer ICSI when sperm have been retrieved surgically or when eggs have been frozen and thawed (1.50.1).
• Consider ICSI if the partner with male reproductive organs has abnormal semen parameters, taking the severity into account, or if a previous IVF cycle resulted in failed fertilisation or a very low fertilisation rate (1.50.2).
• Do not use ICSI for non-male factor fertility problems if the semen parameters are normal (1.50.3).
• Do not use IMSI, intracytoplasmic morphologically selected sperm injection, which picks sperm at very high magnification, as an adjunct to ICSI (1.50.4).
• Do not use PICSI, physiological ICSI, which selects sperm by their binding to hyaluronan, in preference to standard ICSI (1.50.5).

In its rationale, the committee says the evidence it reviewed showed that ICSI for non-male factor problems has no impact on live birth rates, clinical pregnancy rates or other outcomes. It found insufficient evidence that IMSI improves the live birth rate and no difference in live births between PICSI and standard ICSI. It expects the recommendations to reduce the use of ICSI for non-male factor problems, which could save money. IMSI and PICSI are among the extras discussed in IVF add-ons: what the evidence says.

Is it safe?

ICSI has been in clinical use for more than thirty years, and most children born after it are healthy. The HFEA notes that it carries slightly more risks than some other fertility treatments, including a risk that eggs are damaged during preparation and injection. Some genetic and developmental problems have been reported in a very small number of children, but these may be due to the underlying infertility rather than the treatment itself. Where male infertility has a clearly defined genetic cause, especially one on the Y chromosome, sons conceived with ICSI are highly likely to inherit it, and genetic testing and counselling beforehand may be advised.

NICE asks clinicians to tell people that the absolute risks of long-term adverse outcomes of IVF, with or without ICSI, are low, while acknowledging that long-term follow-up is still limited: its research recommendations point out that the longest follow-up in the studies it reviewed was 20 years.

Questions to ask a clinic

• Why is ICSI recommended in our case? Is it because of the semen analysis, a previous fertilisation failure, surgically retrieved sperm or frozen eggs?
• If the semen analysis is normal, what would conventional IVF change for us, and is ICSI charged as an extra?
• Is IMSI or PICSI being offered, and on what evidence?
• If male infertility is severe, is genetic testing advised before treatment?

This article summarises HFEA information, NICE guidance and published research as available on 13 November 2026. It is not personal medical advice: whether ICSI is appropriate depends on tests and history that a fertility specialist can assess.