The influence of treatment and of pretreatment patient characteristics on the probability of long-term disease-free survival in small-cell lung cancer (SCLC) was investigated in a consecutive series of 874 patients. The patients were included in six controlled treatment trials from 1973 to 1981, using different combinations of chemotherapy with or without irradiation. All patients underwent pretreatment staging, including bronchoscopy, peritoneoscopy with liver biopsy, and bone marrow examination. The same procedures were repeated in patients without overt signs of disease 18 months from initiation of treatment, and patients without evidence of SCLC were regarded as long-term survivors. Seventy-two patients were disease-free at restaging, corresponding to 13% of 443 patients with limited-stage disease and 3% of 431 patients with extensive-stage disease. The possible relationship between different pretreatment variables and the probability of 18 months' disease-free survival was investigated by multiple regression analysis. Disease extent was the most important determinant of long-term survival. Being a woman was a positive factor and hypouricemia had negative influence on the long-term results, while features such as performance status and serum lactate dehydrogenase (LDH) did not have significant influence in the regression model. Differences between the efficacy of the applied treatment regimens were less in limited disease than they were in extensive disease, in which six-agent regimens of alternating chemotherapy was significantly better than treatment with three- or four-agent regimens. Accordingly, disease extent seems to be the most pivotal determinant of long-term survival in SCLC, but influence of the patient's sex and serum urate concentration should also be considered.