Uncontrolled studies have suggested that sulphasalazine may be an effective second line agent in rheumatoid arthritis. Sulphasalazine was therefore compared with placebo and intramuscular sodium aurothiomalate in 90 patients with active rheumatoid arthritis. After six months' treatment both sulphasalazine and sodium aurothiomalate had produced significant clinical and laboratory benefit, whereas placebo had produced no significant change in any variable. Thirteen patients stopped taking the placebo because of lack of effect whereas only two patients stopped taking sulphasalazine and one sodium aurothiomalate for this reason. The major toxicity encountered in the group treated with sulphasalazine was nausea or vomiting, or both; this may be related to slow acetylator phenotype. Sulphasalazine appears to be an effective second line agent, and further pharmacokinetic studies might prove useful in diminishing gastrointestinal side effects.