Small cell lung cancer patients who failed primary systemic therapy or who failed after response were randomly assigned to salvage treatment with etoposide (VP-16) and cisplatin (CDDP) or bis-chloro-ethylnitrosourea, thiotepa, vincristine, and cyclophosphamide (BTOC). Good risk patients were those who had tolerated prior chemotherapy well, those who had not had prior radiation therapy, and those who were 65 years of age or younger. Patients with a history of poor tolerance, prior radiation therapy, or those who were older than 65 years of age were classified as poor risk. Forty-five patients were randomized to the BTOC regimen and 58 to the VP-16/CDDP regimen. The overall remission rate was 13% (13 of 103 patients). Good risk patients treated with the BTOC regimen had a remission rate of 27% (three of 11 patients), which was the same rate as patients treated with the VP-16/CDDP regimen (three of 11 patients). Poor risk patients had remission rates of 9% (three of 34 patients) with the BTOC regimen and 9% (four of 47 patients) with the VP-16/CDDP regimen. The median survival time from the start of therapy was 16 weeks for all patients. BTOC good risk patients had a median survival time of 10 weeks, as compared with 14 weeks for poor risk patients. VP-16/CDDP good risk patients had a median survival time of 35 weeks, as compared with 12 weeks for poor risk patients. Although based on small numbers, the advantage in survival time for good risk patients treated with VP-16/CDDP over those treated with BTOC is statistically significant. Prior exposure to VP-16 did not influence the outcome of patients treated with VP-16/CDDP. Both regimens produced moderate toxicity, but were generally well tolerated. It was concluded that VP-16/CDDP may be a useful salvage treatment for good risk patients, despite its limited remission rate. Also, it was found that BTOC has no value for patients in this setting and that neither regimen helps patients who are poor risk.