Objective: To test the hypothesis that liveborn infants conceived by intracytoplasmic sperm injection are at an increased risk of having a major birth defect. Design: Reclassification of the birth defects reported in infants born after intracytoplasmic sperm injection in Belgium and comparison with prevalence estimated in Western Australian population by means of same classification system. Setting and subjects: 420 liveborn infants who were conceived after intracytoplasmic sperm injection in Belgium and 100 454 liveborn infants in Western Australia delivered during the same period. Main outcome measures: Estimates of birth prevalence of birth defects and comparisons of odds ratios between cohort conceived after intracytoplasmic sperm injection and Western Australian infants. Results: Infants born after intracytoplasmic sperm injection were twice as likely as Western Australian infants to have a major birth defect (odds ratio 2.03 (95% confidence interval 1.40 to 2.93); P=0.0002) and nearly 50% more likely to have a minor defect (1.49 (0.48 to 4.66); P=0.49). Secondary data-led analyses, to be interpreted with caution, found an excess of major cardiovascular defects (odds ratio 3.99), genitourinary defects (1.33), and gastrointestinal defects (1.84), in particular cleft palate (5.11) and diaphragmatic hernia (7.73). Conclusions: These results do not confirm the apparently reassuring results published by the Belgian researchers of intracytoplasmic sperm injection. Further research is clearly required. Meanwhile, doctors practising intracytoplasmic sperm injection should bear this alternative interpretation in mind when they counsel couples and obtain informed consent for the procedure. Reports of data from Belgium concluded that there was no increase in the occurrence of major birth defects among infants conceived by intracytoplasmic sperm injection However, the Belgian researchers used a narrow definition of what constituted a major birth defect, and they might have underestimated the comparative prevalence of major defects We reclassified the birth defects in the infants born after intracytoplasmic sperm injection using a standard classification system and compared the results with data from Western Australia that were classified with the same system We found that the infants born after intracytoplasmic sperm injection were twice as likely to have a major birth defect and nearly 50% more likely to have a minor defect Further research is required to elucidate these results, but, meanwhile, they should be born in mind when counselling infertile couples about intracytoplasmic sperm injection