SummaryThe outcome of procedures for assisted reproduction, namely in‐vitro fertilization (IVF) and gamete intrafallopian transfer (GIFT) were used as end‐points to assess the capacity of particular sperm characteristics to predict the fertilizing potential of semen. In addition, the result of IVF performed with spare oocytes from the GIFT procedure, was used to predict the pregnancy outcome of GIFT.The power of sperm characteristics to predict the outcome of assisted reproduction procedures was poor. With regard to the outcome of IVF, sperm morphology was the best discriminant. The threshold value of 16% normal sperm morphology could be used as a good predictor of successful IVF, but could not be applied as a criterion to reject an IVF trial. For the latter purpose, the limit of 5% sperm with normal morphology should be used.The proportion of sperm with grade A motility in native semen provided the best discrimination between GIFT‐fertile and GIFT‐infertile cases. A negative result from IVF performed with spare oocytes was 93% accurate in predicting unsuccessful outcome of GIFT. Hence, diagnostic IVF is the most sensitive assay for male fertility. The occurrence of oocyte fertilization during IVF predicts a successful outcome of GIFT in only 50% of cases. Hence, a negative result in IVF indicates that semen is almost certainly infertile, but a positive result does not guarantee the successful outcome