A double-blind, placebo-controlled trial of efficacy and safety of thalidomide in AIDS-associated wasting was carried out. Ninety-nine of 103 male patients had at least one on-study measurement (intent-to-treat [ITT] cohort). Patients were randomized to thalidomide at 100 mg/day (T100) or 200 mg/day (T200), or placebo for 8 weeks. By ITT analysis, the mean change in body weight of the placebo, T100, and T200 treatment groups was 0.3 kg (0.4%), 2.0 kg (3.0%), and 0.9 kg (1.4%), respectively (p = 0.021 for T100 versus placebo; p = 0.53 for T200 versus placebo). Of the 64 patients who completed the 8 weeks of study treatment, significant weight gain was observed in both the T100 group (2.2 kg, [33%]; p = 0.008 versus placebo) and the T200 group (1.5 kg [2.5%]; p = 0.019 versus placebo). Approximately half the weight gain was fat-free mass (bioimpedance analysis). Patients in the T100 or T200 groups had no significant change in CD4+ cell counts, neutrophil counts, or TNF-alpha levels, compared with placebo. HIV viral load measured as log10 copies/ml decreased by a median of 0.07 in the placebo group, and increased by a median of 0.29 (T100 group) and 0.23 (T200 group) (p = 0.024 andp = 0.018 versus placebo, respectively). Thalidomide therapy was associated with mild to moderate rashes and fevers, but not peripheral neuropathy. Although the anabolic benefits of high-dose thalidomide are limited by drug intolerance, 8 weeks of low-dose thalidomide results in significant weight gain in patients with AIDS-associated wasting.