In a multicenter placebo-controlled double-blind randomized clinical study, 91 patients with rheumatoid arthritis were given 28 days' treatment with recombinant interferon-gamma (50 micrograms daily for 20 days, then 50 micrograms each second day up to day 28, given by subcutaneous injection). The aim of the study was to provide a methodologically clear demonstration of the efficacy of treatment with interferon-gamma, using criteria that could be handled by statistical tests. Evaluatable documentation was available for 79 patients, of whom 40 were treated with the active compound. The principal criterion for the statistical evaluation of the therapeutic success was improvement of the Ritchie "joint pain index" or Lansbury "joint pain index" by at least 30% within 28 days. The chi-square test showed superiority of the interferon arm over the placebo arm with an error probability of alpha less than 1%. In addition, efficacy of interferon-gamma was demonstrated in respect of practically all parameters investigated. The frequency of side-effects, including febrile reactions, was the same for the active compound and the placebo. During interferon treatment the daily maximum body temperature was raised by 0.3 degrees C on average, but was below 37.2 degrees C at all times.