BACKGROUND The clinical Stage I of nonseminomatous germ cell tumors (NSGCT) is inaccurate in 30% of patients. In previous studies on tumor biologic risk factors, low tumor proliferation rates predicted a group of patients at low risk for occult metastatic disease. The goal of this study was to confirm the immunohistochemical assessment of tumor proliferation using MIB-1 (Ki-67 receptor) in a different patient cohort with different investigators to prove the method's reliability. METHODS Orchiectomy specimens of 78 patients with clinical Stage I NSGCT (50 patients with pathologic Stage I and 28 patients with pathologic Stage II disease, all patients underwent retroperitoneal lymph node dissection) were retrospectively analyzed by histopathologic reevaluation and MIB-1 immunostaining. RESULTS Mean MIB-1 values between the two pathologic stages differed significantly (51.5% MIB-1 positive tumors cells in pathologic Stage I and 75.1% MIB-1 positive tumor cells in pathologic Stage II disease; P = 0.02). Using a 70% cutoff value, pathologic stages were correctly classified in 69% of cases (sensitivity of 86%, specificity of 60%, negative predictive value of 88%, and positive predictive value of 55%). Compared with traditional risk factors such as percentage of embryonal carcinoma and vascular invasion, in multivariate analysis, MIB-1 was the best predictor of patients at low risk for metastasis. CONCLUSIONS This study in a different patient population with different investigators confirmed previous results of MIB-1 staining to predict a group of patients with clinical Stage I NSGCT who were at low risk for metastasis. The method is simple and reproducible to improve risk classification in low stage testicular carcinoma. Using this technique, a group of patients at very low risk for metastasis can be identified. [See editorial counterpoint on pages 1641-5 and reply to counterpoint on page 1646, this issue.]