We tested the hypothesis that a combination of sex hormone suppression and inhibition of their target receptors might improve survival for patients with hepatocellular carcinoma (HCC). Eighty-five consequent, previously untreated HCC patients with inoperable disease, were randomized to receive the luteinizing hormone-releasing hormone (LR-RH)-analogue triptorelin and the antiestrogen tamoxifen (33 patients) or triptorelin plus the antiandrogen flutamide (23 patients), or only placebo (29 patients) in a double blind fashion. All groups were comparable as to age, sex, tumor extension, underlying cirrhosis and biochemical parameters. The tamoxifen (TMX) group had a significantly longer survival (282 days) compared with flutamide (112 days) and with placebo (127 days) groups (P = .0238, log rank test). The upper quartile of patients in the TMX group lived 384 days or longer, and most of them (57.1%) were women (P < .0005), in contrast to the upper quartile of the placebo (170 days, 16.7% women) and the flutamide group (134 days, 33.3% women). The calculated tumor volume doubling time (TVDT) was significantly higher in the TMX group (296 days) than in the other two groups (99 and 101 days for placebo and flutamide groups, respectively, P = .023). In a Cox proportional hazards model, the TMX treatment, along with the baseline Okuda's HCC stage, the hepatitis B surface antigen, the portal vein diameter, the carcino embryonic antigen (CEA) and a self-assessment score of quality of life, were covariates predicting survival. Although the degree of serum sex hormone suppression was not a significant predictor of survival, the interaction of female sex and TMX treatment, it was (P = .0052).(ABSTRACT TRUNCATED AT 250 WORDS)