Objective  To assess the feasibility of intermittent hormone therapy for metastatic prostate cancer. Patients and methods  Sixteen patients with metastatic carcinoma of the prostate were treated using a protocol of intermittent hormone therapy with the luteinizing hormone‐releasing hormone (LHRH) analogue leuprorelin at a dose of 3.75 mg every 4 weeks. The protocol required that hormone therapy be stopped when the response was stable, and was designed to assess the duration of the unmaintained response and the probability of a second response to re‐initiation of leuprorelin. Results  Eleven of the 16 patients had a stable hormone response and stopped therapy 3–9 months (mean 5.5) from the start of treatment. The mean (range) duration of the unmaintained response, as judged by monitoring symptoms and serum prostate specific antigen (PSA) levels every month was 4 (2–8) months in the seven patients who had bone metastases and was 3, 3 and 6 months in the three with only loco‐regional disease (one patient temporarily discontinued follow‐up). As the immediate re‐induction of hormone therapy was not mandatory in asymptomatic patients at the time of progression, the mean (range) period off hormone therapy was 8 (3–13) months in those eight patients with bone metastases and was 3, 36 and 42+ months in the three presenting with loco‐regional disease. All 10 patients who re‐initiated hormone therapy had a second hormone response, in six of which led to a decline in PSA level to <2 ng/mL. During the period off hormone therapy no patient developed irreversible symptoms. Conclusions  The temporary cessation of hormone therapy early during the response in patients with metastatic carcinoma of prostate is associated with biochemical evidence of relatively early progression in most cases, but can be associated with significant periods off therapy and with a high chance of a second hormone response. The value of this approach to the quality and duration of patients’ lives requires a prospective comparative evaluation.