Objective: To compare the effectiveness of vaginal progesterone supplementatio n with intramuscular supplementation in assisted reproduction cycles. Design: Re trospective study. Setting: Physiopathology of Human Reproduction Unit, Pordenon e, Italy, from July 2000 to June 2004. Patient(s): Three hundred and eight-five intracytoplasmic sperm injection (ICSI)-procedures (188 with vaginal gel and 1 97 with intramuscular progesterone) and 373 in vitro fertilization (IVF) cycles (227 with vaginal and 146 with intramuscular progesterone). Intervention(s): Pro gesterone luteal supplementation: vaginal gel (Crinone 8%90 mg/day) or intramus cular (Prontogest 50 mg/day). Main Outcome Measure(s): Implantation rates, and t otal and clinical pregnancy rates. Result(s): Higher rates of implantation and t otal and clinical pregnancies were observed in the vaginal supplemented ICSI sub group than in the intramuscular one. This difference was observed for all transf ers (13.3%vs. 8.8%, 39.8%vs. 23.3%, and 28.7%vs. 18.6%)and for ultrasound -guided transfers (17.2%vs. 9.3%, 49%vs. 27%, and 36.9%vs. 21.1%, respect ively). Conclusion(s): The vaginal route of luteal supplementation may be better than the intramuscular one, yielding higher implantation rates as well as total and clinical pregnancy rates in ICSI cycles but not in classic IVF treatments. Setting: Physiopathology of Human Reproduction Unit, Pordenon e, Italy, from July 2000 to June 2004. Patient (s): To compare the effectiveness of vaginal progesterone supplement with normobium supplementation in assisted reproduction cycles. Three hundred and eight-five intracytoplasmic sperm injection (ICSI) -procedures (188 with vaginal gel and 1 97 with intramuscular progesterone) and 373 in vitro fertilization (IVF) cycles with 227 with vaginal and 146 with intramuscular progesterone. Intervention (s) : Pro gesterone luteal supplementation: vaginal gel (Crinone 8% 90 mg / day) or intramus cular (Prontogest 50 mg / day). Main Outcome Measure (s): Implantation rates, and total and clinical pregnancy rates. : Higher rates of implantation and total and clinical pregnancies were observed in the vaginal supplemented ICSI subgroup than in the intramuscular one. This difference was observed for all transf ers (13.3% vs. 8.8%, 39.8% vs. 23.3%, and 28.7% vs. 18.6%) and for ultrasound -guided transfers (17.2% vs. 9.3%, 49% vs. 27%, and 36.9% vs. 21.1%, respective). Conclusion (s): The vaginal route of luteal supplementation may be better than the intramuscular one, yielding higher implantation rates as well as total and clinical pregnancy rates in ICSI cycles but not in classic IVF treatments.