Two hundred eight patients were studied in a prospective, controlled, double-blind multicenter trial comparing auranofin (AUR), gold sodium thiomalate (GST), and placebo. One hundred sixty-one patients completed at least 20 weeks of therapy. Response to a variety of measures of efficacy was generally modest for both gold treatment groups although improvement was continuing in both groups at the end of the study. There was statistically significant improvement with both gold preparations compared to placebo for the number of tender joints, the joint tenderness score, and physician assessment of disease severity. GST was also significantly better than placebo for the joint swelling score. GST demonstrated more improvement in patients with anemia and thrombocytosis compared to the other treatment groups and both gold preparations were superior to placebo in improvement of an elevated erythrocyte sedimentation rate. Twenty-seven percent of patients on GST were withdrawn from the study for adverse drug reaction with rash and stomatitis being the predominant cause. Only 6% of patients on AUR were withdrawn for untoward drug effect. The time of onset of the adverse reactions is discussed. The two gold preparations were similar in efficacy although AUR was better tolerated.