In a 48-week, double-blind trial, 122 patients were randomly assigned to treatment with auranofin (60) and gold sodium thiomalate (GST) (62) at five centers. Both groups showed significant improvement (P less than 0.05) from baseline in parameters of disease activity. Results of the covariance analysis for all patients who completed the trial showed no significant differences (P less than 0.05) in efficacy between the two groups. The proportions of patients showing 50% or greater improvement in tender joints, swollen joints, activity index, severity of pain, general health rating, and erythrocyte sedimentation rate (ESR) were similar for both auranofin-treated and GST-treated patients who completed the 48-week trial. When all patients who entered the trial were evaluated, a slightly greater proportion of patients on auranofin had improved. Diarrhea occurred more frequently with auranofin (32%) compared to GST (19%), whereas rash and pruritus were twice as common in those patients treated with GST compared to those treated with auranofin. The withdrawal rate due to adverse reactions was 10% for auranofin vs 26% for GST. It was concluded that the efficacy of auranofin was comparable to that of injectable gold and was better tolerated, as evidenced by the lower withdrawal rate from adverse events for the auranofin patients.