In an unselected group of 278 patients with germ cell tumors, disease-free status was obtained in 97% by a treatment program including a surveillance-only strategy for stage I testicular cancer, and low-or high-dose cisplatinum-etoposide treatment for patients with more extensive disease. The overall follow-up period was a median of 40 months (range 20-62 months). At present 100% of patients with stage I disease, 91% with stage II disease, 86% with stage III disease, 75% with extragonadal germ cell tumors, and 3 of 3 patients with germ cell tumors in the ovary are alive and without disease. Among 36 patients treated with high-dose cisplatinum and etoposide there were six toxic deaths, four of them in patients with residual malignant disease. Three patients died of progressive disease. There were no toxic deaths among 54 patients with disseminated disease but without poor prognostic features who were treated with low-dose cisplatinum-etoposide; six of these patients died of progressive disease. It is concluded: 1) that surveillance is a feasible and reasonable strategy for patients with stage I disease; 2) that excellent survival results can be achieved with standard-dose cisplatinum-etoposide in patients with disseminated disease and a favorable prognostic profile; and 3) that disease-free status can be obtained in nearly all patients with poor prognostic features at the expense of significant toxicity. Standardized criteria for selection of patients with poor prognoses are needed. Randomized trials should be carried out to define the role of high-intensity treatment and, finally, measures to decrease or prevent serious toxicity should be explored.