The therapeutic benefit of chemotherapy in androgen independent prostate cancer is limited. 5-Fluorouracil has been reported to have modest antitumor activity in androgen independent prostate cancer. Although alpha-interferon is inactive as a single agent in prostate cancer, preclinical data indicate that it increases the in vitro cytotoxicity of 5-fluorouracil against a variety of malignant cells. We evaluated the relative antitumor activity and tolerance of 5-fluorouracil versus 5-fluorouracil plus alpha-interferon in 50 patients with histologically confirmed metastatic adenocarcinoma of the prostate. These patients had progressive disease in the presence of castrate levels of testosterone. A prospective randomized phase II open labeled trial was performed because of the difficulty in measuring responses in patients with metastatic prostate cancer. Of 23 patients treated with 5-fluorouracil alone and 28 treated with 5-fluorouracil plus alpha-interferon 17 and 23, respectively, were evaluable for response and toxicity, and 5 and 5, respectively, were evaluable for toxicity only. Only 2 of 17 (11.7%) and 4 of 23 (17%) patients, respectively, showed a greater than 50% decrease in serum prostate specific antigen (no significant difference). There was no difference in duration of response or duration of survival between the 2 groups (mean duration of response 8.64 and 6.17 weeks, respectively, and mean duration of survival 33.70 and 38.65 weeks, respectively). Both regimens caused significant morbidity (mucositis and neurotoxicity) and 3 treatment related deaths at the high 5-fluorouracil doses. 5-Fluorouracil alone and with alpha-interferon at the doses used have minimal antitumor activity against androgen independent prostate cancer and, therefore, should not be tested further in these patients. Androgen independent prostate cancer selected using our criteria is a rapidly progressive disease, and these patients are an ideal target population for phase II studies.