PURPOSE: To clarify the controversy about the management and prognosis of human chorionic gonadotropin-producing seminoma, the records of 132 patients with abnormal human chorionic gonadotropin values treated with radiotherapy were analyzed. METHODS AND MATERIALS: The records of 1169 patients with pure seminoma treated in 10 institutions were screened for serum or urinary human chorionic gonadotropin. One hundred and thirty two patients with elevated human chorionic gonadotropin were found: 96 Stage I, 20 IIA, 7 IIB, 8 III and 1 IV. Median age was 34 y., mean follow-up was 5.0 years [range 1-12 y]. All received infradiaphragmatic radiotherapy (median dose 30 Gy), 25 (2 Stage I, 11 IIA, 5 IIB and 7 III) supradiaphragmatic radiotherapy (median dose: 28.5 Gy) and 10 had also initial chemotherapy (3 Stage IIB 6 III and 1 IV). Patients were allocated to three groups according to human chorionic gonadotropin values: (a) moderate elevation: up to 10 times (104 pts), (b) high elevation: 10 to 100 times (20 pts), (c) very high elevation: over 100 times the upper limit of normal value (8 pts). RESULTS: The proportion of Stage I, II and III was 76%, 19%, 5% in the ME group versus 50%, 35%, 15% in the high elevation group (p < 0.05). In the very high elevation group there were 7 Stage I and 1 Stage IV. Of 132 patients, six died (three dead of disease, two suicides, one acquired immunodeficiency syndrome). The 5 years overall survival probability was 94%. There were seven recurrences (initial stage: 1 Stage I, 2 IIB, 3 III and 1 IV). Of these, there were one in-field recurrence, 3 out of field and 3 in both sites. In 5 of 7, the human chorionic gonadotrophin level was again elevated at recurrence. The 5 years recurrence-free-survival probability was 94% (98% for Stage I, 100% for Stage IIA and 65% for Stage IIB and III [p < 0.001 between I and IIB + III, p < 0.05 between IIA and IIB + III]). Four of the 7 recurrences were salvaged by chimiotherapy +/- radiotherapy. In the high elevation and very high elevation groups, the 5 years recurrence-free-survival was 88%, vs. 96% for the moderate elevation group (p = 0.10). CONCLUSION: Based on this series of patients, human chorionic gonadotropin production is not an unfavorable prognostic factor in pure seminoma. Even in the subgroups with high or very high human chorionic gonadotropin levels (who had a higher proportion of advanced stages), the prognosis remained excellent. In Stage I and IIA seminoma with abnormal human chorionic gonadotropin levels, recurrence rate after post-operative radiotherapy alone is extremely low.