To evaluate the clinical efficacy of interferon-alpha in hepatocellular carcinoma, 71 adult Chinese patients with histologically proven inoperable hepatocellular carcinoma were randomized to receive recombinant interferon-alpha 2a (50 x 10(6) IU/m2) intramuscularly three times a week (n = 35) or no antitumor therapy (n = 36). The survival of interferon-alpha-treated patients was significantly better than that of patients who received no antitumor therapy (p = 0.0471); median lengths of survival were 14.5 and 7.5 wk, respectively. Objective tumor regression greater than 50% was observed in 31.4% (11 of 35) of patients receiving interferon-alpha. Interferon-alpha induced tumor regression greater than 50% in 11 (31.4%) patients. Compared with the group receiving no antitumor therapy, the interferon-alpha therapy group had more tumor regression (p < 0.0001) and less tumor progression (p = 0.001). This high-dose interferon-alpha therapy was relatively well tolerated; only 34.3% of patients required reduction of dosage by one third or one half because of persistent fatigue. Two patients with diabetes mellitus (one also had tabes dorsalis) exhibited mental deterioration that might have been partially attributable to interferon-alpha therapy. We conclude that interferon-alpha is useful in a proportion of Chinese patients with inoperable hepatocellular carcinoma, both in prolonging survival and in inducing tumor regression.