In a 2-year, randomized, double-blind Nordic multicentre trial, auranofin was compared with placebo in early (disease duration less than or equal to 2 years), active rheumatoid arthritis (RA). Efficacy and safety were analysed in 67 patients receiving auranofin and 65 receiving placebo. Life table analysis demonstrated a significantly higher withdrawal rate due to insufficient therapeutic effect in the placebo group, whereas more patients dropped out due to side effects in the auranofin group. More auranofin than placebo patients (35 vs. 24) completed the 2 years. Clinical and inflammatory activity improved in both groups, but consistently more so in the auranofin group, in spite of the greater consumption of local steroids and NSAIDs in the placebo group. The most frequent side effects leading to withdrawal in the auranofin group were cutaneous and gastrointestinal reactions. The study demonstrated that most patients exhibit improvement in clinical signs and symptoms and about half of all patients with early RA continue to take auranofin for at least 2 years.