We previously reported the results of a pilot study of intermittent androgen deprivation (IAD) therapy in which surveillance was performed when PSA level fell below 0.3 ng/ml and androgen deprivation was resumed when PSA level exceeded 2.0 ng/ml. In the present study, we compared the duration of androgen dependence in patients treated with IAD with that in patients with continuous androgen deprivation (CAD) therapy. Forty-six patients with clinically localized or metastatic prostate cancer, or biochemical recurrence after radical prostatectomy were treated with IAD from 1995 to 2003. Patients in or after the second cycle of IAD (30 patients) were evaluated for duration of androgen dependence. Patients whose serum PSA nadir became <0.3 ng/ml (33 patients) represented a control group of CAD. The overall 5-year biochemical progression-free rate was 58% and 89% in the IAD and CAD groups, respectively; there was no significant difference between the two groups (p=0.5). Subgroup analysis showed that, irrespective of metastasis, the 5-year biochemical progression-free survival rate in the IAD group was not significantly different from that in the CAD group. However, IAD offered significantly better results for well-differentiated prostate cancer, whereas CAD offered significantly better results for moderately or poorly differentiated prostate cancer. The results obtained from this retrospective and nonrandomized study suggested that IAD may be a feasible treatment for well-differentiated prostate cancer.