The effect of the testosterone derivative oxymetholone alone or in combination with the H1-receptor antagonist ketotifen, which has recently been shown to block tumour necrosis factor alpha (TNF alpha), on weight gain and performance status in human immunodeficiency virus (HIV) patients with chronic cachexia was evaluated in a 30-week prospective pilot study. Thirty patients were randomly assigned to either oxymetholone monotherapy (n 14) or oxymetholone plus ketotifen (n 16). Patients receiving treatment were compared with a group of thirty untreated matched controls, who met the same inclusion criteria. Body weight and the Karnofsky index, which assesses the ability to perform activities of daily life, and several quality-of-life variables were measured to evaluate response to therapy. The average weight gain at peak was 8.2 (SD 6.2) kg (+ 14.5% of body weight at study entry) in the oxymetholone group (P < 0.001), and 6.1 (SD 4.6) kg (+10.9%) in the combination group (P < 0.005), compared with an average weight loss of 1.8 (SD 0.7) kg in the untreated controls. The mean time to peak weight was 19.6 weeks in the monotherapy group and 20.8 weeks in the combination group. The Karnofsky index improved equally in both groups from 56% before to 67% after 20 weeks of treatment (P < 0.05). The quality of life variables (activities of daily life, and appetite/nutrition) improved in 68% (P < 0.05) and 91% (P < 0.01) of the treated patients respectively. Oxymetholone was safe and promoted weight gain in cachectic patients with advanced HIV-1 infection. The addition of ketotifen did not further support weight gain. These results suggest the need for a randomized, double-blind, placebo-controlled multicentre trial.