Methotrexate appears to be an effective alternative to corticosteroid therapy for some patients with sarcoidosis. The mechanism of action of methotrexate as an immunosuppressive is unknown. Patients with symptomatic pulmonary sarcoidosis underwent pulmonary function tests and bronchoscopy with bronchoalveolar lavage. Patients were treated with 10 mg methotrexate or prednisone weekly for at least 6 months and repeat studies were performed. A comparison was made between those patients receiving methotrexate (12 patients) and those receiving prednisone (12 patients). For both groups, there was a significant improvement in the vital capacity with therapy (Prednisone: Pre = 2.5 +/- 0.14 L (Mean +/- SEM); Post = 3.1 +/- 0.18 L, p less than 0.01; Methotrexate: Pre = 2.4 +/- 0.14 L; Post = 2.8 +/- 0.18 L, p less than 0.01). In addition, the percentage of lymphocytes in the lavage fell significantly for both the prednisone (Pre: 30 +/- 3.5%; Post: 16 +/- 2.7%, p less than 0.001) and methotrexate (Pre: 37 +/- 3.4%; Post: 13 +/- 2.9%, p less than 0.001) groups. Alveolar macrophages from the symptomatic sarcoid patients were found to be spontaneously releasing hydrogen peroxide and tumor necrosis factor. After treatment with either prednisone or methotrexate, alveolar macrophages retrieved by lavage spontaneously released less of either macrophage product. We found that effective doses of methotrexate for sarcoidosis led to significant changes in lymphocyte and macrophages retrieved by lavage.