Although the majority of patients with disseminated germ cell tumours can be cured with cisplatin-based chemotherapy, mortality is still up to 20%. Several prognostic factors have been identified to differentiate between patients with a good, intermediate or poor prognosis. In this review we discuss the recent chemotherapy trials, which were designed to reduce toxicity in good-prognosis patients and to improve efficacy in intermediate- and poor-prognosis patients. In good-prognosis patients it is obvious that the omission of bleomycin and the replacement of cisplatin by carboplatin has no place in first-line standard treatment. The reduction of four standard courses of bleomycin, etoposide and cisplatin (BEP) to three is shown possible in one study, but a confirmatory study is currently ongoing in the EORTC. In intermediate- and poor-prognosis patients, the use of new agents or alternating regimens (with or without shortened intervals) did, by now, not improve final outcome. The role of high-dose chemotherapy remains to be determined. Against this background, four courses of standard-dose BEP should still be considered treatment of first choice in the majority of patients with disseminated germ cell tumours. Furthermore, the policy in clinical stage I disease has been reviewed. In clinical stage I seminoma patients the policy is to apply adjuvant radiotherapy, while the strategy in patients with non-seminomatous tumours (surveillance, retroperitoneal lymph node dissection or adjuvant chemotherapy in high-risk patients) depends highly on the local situation, such as the operating skills of the urologist, and on the possibilities for tight follow-up. Of patients with true resistance for up-front BEP chemotherapy 90% will normally die. In patients who achieve a complete response on first-line chemotherapy, but relapse thereafter 30% will have no evidence of disease with second-line chemotherapy (VIP). In this group of patients results with high-dose chemotherapy seem promising, but its value should preferentially be determined in either a randomized fashion or by long-term follow-up from a large group of patients according to a similar protocol. The use of post-chemotherapy surgery is an essential part of management for metastatic non-seminomatous germ cell tumours, while the majority of residual masses in pure seminoma will disappear spontaneously, and frequent follow-up is recommended instead of surgical intervention.