To assess the relative efficacy of methotrexate (MTX), azathioprine (AZA), and their combination in the treatment of rheumatoid arthritis (RA), a double blind, prospective, multicenter, controlled trial was carried out. Two hundred nine patients with active RA were treated with escalating doses of MTX (5-15 mg/week), AZA (50-150 mg/day), or combination (5/50-7.5/100) with opportunity to increase dosage at 6 week intervals. Patients were evaluated for clinical and laboratory improvement and assessed for radiologic progression at 48 weeks. One hundred ten patients remained on the initial, randomly assigned therapy. Response was defined as 30% or greater improvement in at least 3 of 4 variables, and occurred in the following: 38% on the combination arm, 26% on AZA, and 45% on MTX (p = 0.06). A trend for decreased radiologic progression was seen in the MTX group. Adverse experiences and treatment termination occurred more frequently in the combination and AZA arms relative to the MTX group. The most frequent causes for treatment discontinuations were lack of effectiveness, gastrointestinal adverse effects and liver enzyme elevation. This study establishes that the combination of MTX and AZA in the dosages employed is not associated with more toxicity than treatment with single agents, but enhanced efficacy is not seen. A trend toward decreased radiographic progression was noted in the MTX treated patients.