In patients with obstructive azoospermia in whom standard microsurgical procedures fail or are unfeasible, the only source of spermatozoa is the testicle. In addition, in some azoospermic patients with severe spermatogenic failure, a few spermatozoa may be present in testicular biopsy specimens despite high serum follicle stimulating hormone concentrations. In all these cases, intracytoplasmic sperm injection (ICSI) with testicular biopsy-extracted spermatozoa may offer the chance of pregnancy. To assess the efficacy of this procedure, we compared the results of two series of ICSI cycles performed during the same time period: 21 cycles using testicular biopsy-extracted spermatozoa and 83 cycles using ejaculated spermatozoa. Mean fertilization rates (59% with testicular and 68% with ejaculated spermatozoa), mean cleavage rates (93% with testicular and 90% with ejaculated spermatozoa), embryo quality (77% good quality embryos in the testicular sperm group and 77% in the ejaculated sperm group) and clinical pregnancy rates (36.8% in the testicular sperm group and 28% in the ejaculated sperm group) were not significantly different in both groups. We conclude that high fertilization, cleavage and pregnancy rates can be achieved with intracytoplasmic testicular sperm injection, reaching levels comparable with those of ICSI using ejaculated spermatozoa.