In routine semen analysis, discrepancies may occur between sperm acrosin activity test results and sperm acrosomal morphology.Discrepant test results of sperm acrosin activity (spectrophotometric assay) vs. sperm morphology (strict criteria) in the initial diagnostic investigation of 107 infertile couples were evaluated with respect to fertilization rate (% oocytes with 2 pronuclei) further obtained in IVF treatment.Acrosin activity positively correlated with sperm morphology (% normal forms) (r=0.537) and fertilization rate (r=0.526). ROC curves for the prediction ofor =50% fertilization rate were comparable for acrosin activity and sperm morphology, with optimal cutoff values at 25 microIU/10(6) sperm and 10%, respectively. In multiple regression analysis, sperm acrosin activity (P=0.002) predicted fertilization rate independently of sperm morphology (P0.001) and sperm vitality (eosin-nigrosin stain) (P=0.03). Acrosin activitiesor =25 microIU/10(6) sperm were observed in 36% of severe teratozoospermic samples (or =4% normal spermatozoa) associated with low fertilization rate. Twenty percent of the morphologically normal ejaculates showed a low acrosin activity (25 microIU/10(6) sperm) and low hypoosmotic swelling test (HOST) scores (31.4+/-7.6%) and were associated with low fertilization rate.The sperm acrosin assay can help to predict sperm fertilizing capacity in IVF independently of sperm morphology.