A double-blind comparison of the clinical efficacy and tolerance of varying doses of diflunisal (DFS) and acetylsalicylic acid (ASA) was carried out in 15 patients with rheumatoid arthritis who were given no other antirheumatic medication. An effort was made to select appropriate anamnestic, functional and sociofunctional tests and to optimize their validity by careful measurements performed by the same specialized physiotherapist and occupational therapist. In addition, the serum concentration of DFS and salicylic acid were monitored by high-pressure liquid chromatography. The therapeutic effects of DFS was at least as good as that of ASA. Moreover, DFS was better tolerated; all 7 patients on DFS could sustain the maximum dose (1g) of this drug, while that of ASA (4g) was tolerated by only one of 8 patients. All these experienced side effects, necessitating drug withdrawal in 3 cases, one being a serious hepatotoxic reaction. DFS treatment, on the other hand, was associated with only one minor side effect. The side effect difference was statistically significant (p less than 0.01). The analyses of drug concentrations in serum verified that all patients were exposed to DFS and ASA as planned, adding safety to the judgement of the therapeutic effects. The findings support the view that the novel salicylic acid derivative DFS may offer a therapeutic advantage in the treatment of rheumatoid arthritis; it seems to have at least the same therapeutic effect as ASA an may be better tolerated.