The efficacy and tolerability of aceclofenac, a new non-steroidal anti-inflammatory drug (NSAID) was compared to that of ketoprofen in a multicentre, double-blind, prospective, randomized study of 3-months duration in patients with rheumatoid arthritis. One hundred and sixty-nine patients, aged between 22 and 70 years, were included in the study. Patients were randomly assigned to two treatment groups; either aceclofenac 100 mg b.i.d. (87 patients) or ketoprofen 50 mg t.i.d. (82 patients). A placebo tablet was administered to aceclofenac-treated patients to maintain the double-blind conditions of the study. Patients were examined at 15 days and at 1, 2 and 3 months. Efficacy was assessed by the following clinical parameters: Ritchie index, pain on a visual analogue scale, grip strength, morning stiffness, spontaneous morning pain, pain on movement and nocturnal pain, together with functional capacity. Efficacy was demonstrated for both drugs, with progressive improvement in the main clinical evaluation parameters until the end of the treatment period. This was particularly pronounced at 15 days in the aceclofenac group, with a rapid improvement in the Ritchie index (baseline vs 15 days: P < 0.001). Laboratory analyses performed were all within the normal range for both drugs. Eleven patients in the ketoprofen group abandoned the study because of inefficacy, whilst only 4 patients discontinued the treatment for this reason in the aceclofenac group. Eleven patients in the ketoprofen group and 2 patients in the aceclofenac group withdrew from the study because of adverse events. In summary, this study demonstrated that aceclofenac, a new NSAID, is effective in the symptomatic treatment of RA with a minor number of patient withdrawals because of inefficacy and a better safety profile than ketoprofen.