Aim: To evaluate the fertilization competence of spermatozoa from ejaculates and testicle when the oocytes were matured in vitro following intracytoplasmic sperm injection (ICSI). Methods: Fifty-six completed cycles in 46 women with polycystic ovarian syndrome were grouped according to the semen parameters of their male partners. Group 1 was 47 cycles that presented motile and normal morphology spermatozoa in ejaculates and Group 2 was the other nine cycles where male partners were diagnosed as obstructive azoospermia and spermatozoa could only be found in testicular tissue fragment. All female patients received minimal stimulation with gonadotropin. Immature oocytes were matured in vitro and inseminated by ICSI. The spermatozoa from testes were retrieved by testicular fine needle aspiration. Results: A total of 449 and 78 immature oocytes were collected and cultured for 48 hours, 75.5 % (339/449) and 84.6 % (66/78) oocytes were matured in Groups 1 and 2, respectively. The percentage of oocytes achieving normal fertilization was significantly higher in Group 1 than that in Group 2 (72.9 % vs. 54.5 %, P < 0.05). There were no significant differences in the rates of oocytes cleavage and clinical pregnancies in these two groups [87.4 % (216/247) vs. 88.9 % (32/36); 21.3 % (10/47) vs. 44.4 % (4/9)]. A total of 15 babies in the two groups were healthy delivered at term. Conclusion: It appears that IVM combined with ICSI using testicular spermatozoa can produce healthy infants, while the normal fertilization rate of in vitro matured oocytes after ICSI using testicular spermatozoa was significantly lower than using the ejaculated spermatozoa. Aim: To evaluate the fertilization competence of spermatozoa from ejaculates and testicle when the oocytes were matured in vitro following intracytoplasmic sperm injection (ICSI). Methods: Fifty-six completed cycles in 46 women with polycystic ovarian syndrome were grouped according to the semen parameters of Their male partners. Group 1 was 47 cycles that presented motile and normal morphology spermatozoa in ejaculates and Group 2 was the other nine cycles where male partners were diagnosed as obstructive azoospermia and spermatozoa could only be found in testicular tissue fragments. All female patients received minimal stimulation with gonadotropin. Immature oocytes were matured in vitro and inseminated by ICSI. The spermatozoa from testes were retrieved by testicular fine needle aspiration. Results: A total of 449 and 78 immature oocytes were collected and cultured for 48 hours, 75.5% (339 / 449) and 84.6% (66/78) oocytes were matured in Groups 1 and 2, respectively. The percentage o F oocytes achieving normal fertilization was significantly higher in Group 1 than that in Group 2 (72.9% vs. 54.5%, P <0.05). There were no significant differences in the rates of oocytes cleavage and clinical pregnancies in these two groups [87.4% (216/247) vs. 88.9% (32/36); 21.3% (10/47) vs. 44.4% (4/9)]. A total of 15 babies in the two groups were healthy delivered at term. Conclusion: It appears that IVM combined with ICSI using testicular spermatozoa can produce healthy infants, while the normal fertilization rate in vitro matured oocytes after ICSI using testicular spermatozoa was significantly lower than using the ejaculated spermatozoa.