Between April 1980 and December 1985 a prospective controlled and randomized multicenter study of 124 assessable patients with stage I and II epithelial ovarian carcinomas was carried out. The aims of this study were first to verify the value of adjuvant irradiation therapy or combined chemo-irradiation therapy, and second to evaluate the importance of different prognostic factors such as age, histology, tumor grading, and tumor stage. Patients with well-differentiated stage IA tumors did not receive any therapy; patients with undifferentiated stage IA tumors were randomized to "no therapy" or irradiation therapy; patients with stage IB, IIA, and IIB tumors were treated by either irradiation or a combined chemo-irradiation therapy consisting of Adriamycin/Cyclophosphamide. In patients with stage IC and IIC ovarian carcinomas a combined irradiation-polychemotherapy was instituted, consisting either of Adriamycin/Cyclophosphamide or of Adriamycin/Cisplatin. Because of the low number of patients and the relatively good prognosis no definite evaluation of the individual therapeutic modalities could be done. An analysis of all patients showed that differentiation between stage I and stage II ovarian carcinomas is the single most important prognostic factor. After a mean observation time of 42 months (8-71 months) for stage I tumors a 91% probability of three-year survival was attained, in comparison with 57% for stage II tumors. It has not been definitely established that there is a need for adjuvant therapy for stage I tumors and this should be confirmed by further studies. Because of the unfavorable prognosis for patients with stage II ovarian cancer, these patients should be given the same polychemotherapy, including Cisplatin, as patients with advanced ovarian carcinomas.