Twenty-seven newly diagnosed patients with idiopathic pulmonary fibrosis (IPF) who were previously untreated for IPF were enrolled in a prospective, double-blind, randomized, placebo-controlled study to compare the therapeutic effect of combined prednisone/azathioprine (n = 14) with prednisone plus placebo (n = 13). Prednisone was started at 1.5 mg/kg/day (not to exceed 100 mg/day) for the first 2 wk followed by a biweekly taper to a maintenance dose of 20 mg/day. Azathioprine was administered at a daily dose of 3 mg/kg (not to exceed 200 mg/day). The patients tolerated the use of azathioprine well with few associated side effects. Changes in lung function at 1 yr, as measured by resting alveolar-arterial oxygen difference P[A-a]O2, FVC, and single breath diffusing capacity for carbon monoxide (DLCOSB), were all somewhat better in the azathioprine/prednisone group compared with the prednisone alone group, although none of these comparisons were statistically significant. Six of 14 (43%) patients randomized to prednisone plus azathioprine died during the 9-yr follow-up period, compared with 10 of 13 (77%) patients randomized to prednisone plus placebo. A Cox model survival analysis shows a nonsignificant but potentially large survival advantage for azathioprine/prednisone (hazard ratio 0.48, with 95% confidence interval increasing from 0.17 to 1.38). When adjusted for age, the survival advantage of azathioprine/prednisone becomes marginally significant (hazard ratio 0.26, with 95% confidence interval increasing from 0.08 to 0.88; p = 0.02 by large sample approximation, p = 0.05 by randomization test). We conclude that combined prednisone and azathioprine is a safe and possibly effective regimen for the treatment of IPF.(ABSTRACT TRUNCATED AT 250 WORDS)