Fifty-two patients with rheumatoid arthritis were studied in a single blind trial comparing aurothioglucose and auranofin. The duration of the study was 52 weeks. Twenty-six patients, 13 in each treatment group, dropped out during the first year of treatment. The main reason for discontinuing treatment with aurothioglucose was adverse reactions and, in the auranofin group, lack of efficacy. In those patients who continued therapy the results of treatment were comparable; patients on aurothioglucose improved slightly more than auranofin treated patients, although the difference was not statistically significant.