The aim of our study was to compare the outcome of intracytoplasmic sperm injection (ICSI) with fresh and frozen-thawed epididymal spermatozoa retrieved by percutaneous epididymal sperm aspiration (PESA) or microepididymal sperm aspiration (MESA) from patients with obstructive azoospermia. A retrospective analysis of consecutive ICSI cycles was performed, comparing the outcome in 24 patients with obstructive azoospermia undergoing surgical sperm aspiration by MESA (7 cycles) or PESA (17 cycles). In 23 of 24 patients, excess spermatozoa were cryopreserved. Following thawing, 21 ICSI cycles were performed (11 cycles after MESA, 10 after PESA). No statistically significant differences were noted in all parameters examined in ICSI cycles with fresh or cryopreserved spermatozoa from the same patients. Comparing all ICSI cycles with fresh and frozen-thawed epididymal spermatozoa, the rates of two-pronuclear fertilization (56% versus 53%), embryo cleavage (90% versus 86%), implantation (10% versus 14%), clinical pregnancy per embryo transfer (32% versus 37%) and delivery/ongoing pregnancy rate (27% versus 26%) were not statistically different. The cumulative ongoing pregnancy rate per sperm retrieval procedure was 46%, respectively. We conclude that the clinical outcome of ICSI with fresh and frozen-thawed spermatozoa after retrieval by PESA was similar to that by MESA. Epididymal sperm cryopreservation in patients with obstructive azoospermia is feasible and efficient using a simple freezing protocol and should be offered to optimize the yield of pregnancies achieved following such procedures.