In 1982 there were four reports of cisplatin-induced remissions of invasive thymoma. These observations led the Eastern Cooperative Oncology Group (ECOG) to conduct a Phase II trial of cisplatin (50 mg/m2 intravenously every 3 weeks). During a 4-year period 24 patients were entered on this trial; 3 were excluded because review of histologic material failed to confirm the presence of thymoma. The characteristics of the remaining 21 patients were as follows: median age, 51; males/females, 11/10; ECOG performance status, 0-1/2-3, 16/5; previous treatment with chemotherapy, 3; previous surgery, 20; previous radiation, 15; weight loss, < 5%/ > or = 5%: 17/4. One patient was eliminated from response analysis because of failure to return for follow-up tumor measurements. The following responses were observed in the remaining 20 patients. Partial remission, 2 (10%); stable disease, 8 (40%); and progression, 10 (50%). The median survival was 76 weeks, and the 2-year survival rate was 39%. Four patients experienced severe nausea and vomiting, but life-threatening and lethal toxicities were not observed. Cisplatin given at this dose was relatively ineffective in producing tumor regression in recurrent or metastatic thymoma.