Risk factor analysis to identify low-risk patients for occult metastatic disease (vascular invasion, percentage embryonal carcinoma, MIB-I proliferation rate) yields reliable results if performed by experts. A correct prediction is possible at the 90% level. Similar accuracy, however, may be achieved if the computed tomography (CT) staging is optimized and the evaluation performed by an experienced investigator. The combination of both methods (biological risk factor analysis and CT staging) may virtually exclude the risk of relapse in a limited number of patients. However, so far, no risk factor that is able to reliably predict occult metastatic disease or relapse in clinical state I patients has been identified in prospective trials. The preliminary results of the current German Multicenter Trial suggest an inferior value of prediction for low-risk patients if risk factor analysis and/or CT staging is performed in non-specialized centers.