To assess the effect of megestrol acetate (MA) on the appetite and weight of cancer patients with nonhormone-dependent tumors, a double-blind, placebo-controlled trial was designed. One hundred fifty patients were included: 76 were given MA (240 mg/day orally) for at least 2 months, and 74 were given placebo (P). Body weight, subjective sense of appetite (SSA) evaluated by an analogic linear visual scale scored from 1 to 10, and performance status (PS) were measured before therapy and monthly thereafter. No differences in body weight before and after treatment could be found in any group. However, 32% of the patients in the MA group (95% confidence interval: 20.1-44.6%) gained 2 or more kilograms (P less than 0.001). This group also showed an improvement in SSA (P less than 0.01): an increase greater than or equal to 2 points appeared in 57.5% of patients (95% confidence interval 44.6-69.4%). There was no significant difference in PS for the treatment groups before or after therapy. The percentage of reported adverse events did not differ significantly from one treatment group to the next. We conclude that therapy with MA at a dose of 240 mg/day improved SSA and was associated with moderate weight gain in patients with hormone-insensitive malignancies.