Placebo and chlorambucil (0.05-0.1 mg/kg/day) were compared as treatments for scleroderma in a 3-year, randomized, double-blind, parallel study of 65 patients. Slopes of change over time, calculated for each organ system of each patient, failed to show significant differences between treatment groups. An immunosuppressive effect in the chlorambucil-treated group was shown by decreases in the white blood cell counts (P = 0.02), platelet counts (P = 0.04), lymphocyte counts (P = 0.001), IgA/IgM concentrations (P less than or equal to 0.05), and wheal size on mumps skin tests (P = 0.02). Three years of immunosuppressive treatment with chlorambucil did not benefit this scleroderma population.