From May 1981 through December 1990, 167 patients with ovarian carcinoma were treated with whole abdominal irradiation. All patients of the first group (n = 62) presented with no or minimal (< 2 cm) residual disease at the time of irradiation. They received whole abdominal irradiation after initial surgery. Residual disease was left after initial surgery for patients of the second group (n = 105). They underwent a cisplatin polychemotherapy regimen followed by second-look laparotomy and abdominopelvic irradiation. The irradiation was performed by a four orthogonal field's technique. Thirty grays were given with a 25 MV photon beam. Boosts were performed in 50 cases (mean dose of 15 Gy); 91% have completed irradiation at the prescribed dose level; 5% had severe intestinal toxicity requiring surgery. The median follow-up was 68 months. The 5-year actuarial survival rate was 50%, 67% group 1.40% group 2. Five-year actuarial survival was 84% for T1, 61.5% for T2, 38% for T3. Five-year survival depended on tumor rest after initial surgery (residual disease 36.5%, no residual disease 70%). For patients in group 2, the size of residual disease at second-look laparotomy was a significant prognostic factor: five-year actuarial survival 76% with negative second-look, microscopic residual disease (positive cytology or random biopsies) 60%, microscopic disease after cytoreductive surgery at second-look 22%, macroscopic residual disease (small nodules) 10%. These results indicate that abdominopelvic irradiation is valuable in the treatment of ovarian carcinoma when there is no macroscopic residual disease after initial surgery or second look laparotomy. To assess the place of radiotherapy, randomized trials comparing whole abdominal irradiation with chemotherapy, watch policy for staged patients is warranted.