We report here in the interim analysis of an ongoing randomized clinical trial designed to test whether androgen priming enhances tumor chemosensitivity in men with stage D prostate cancer refractory to orchiectomy. All patients are continuously treated with aminoglutethimide and hydrocortisone, to lower adrenal androgen secretion, and are given cyclic chemotherapy. Patients in the stimulation arm receive also the synthetic androgen, fluoxymesterone, for 3 days before and on the day of chemotherapy. Of 57 patients entered to date, 41 have received adequate treatment to be evaluable. Response to therapy (objective remissions + stabilizations of disease) occurred in 17 of 18 evaluable patients (94%) randomized to the stimulation arm, and in 16 of 23 evaluable patients (70%) in the control group (p less than 0.025). Duration of response was not significantly different in the two groups (median: 9 months in the stimulation and 12 months in the control arm). With 30% of the total of 57 patients still alive, survival is not significantly different in the stimulation (median: 13 months) and control arm (median: 16 months). As expected, patients who responded to treatment lived significantly longer than those who failed to benefit. Two episodes of reversible spinal cord compression occurred during androgen administration. (The risk of this serious side effect may be reduced by performing a screening myelogram to rule out subclinical spinal metastasis). Our preliminary data suggest that androgen priming may enhance the tumoricidal effect of cytotoxic drugs in advanced prostate cancer. The lack of improvement in duration of response and survival may be explained by the large fraction of hormone-independent cells probably present in patients with tumors refractory to orchiectomy.