BACKGROUND: The primary aim was to induce a high number of pCR in early (FIGO IC, IIB + C) - and advanced (FIGO III-IV) - stage ovarian cancer with a surgery plus 4 cycles of cisplatin and melphalan (PAMP) regimen. The second objective was to prevent relapse with WAR in patients in remission after chemotherapy. PATIENTS AND METHODS: 218 eligible patients were treated after staging laparotomy with cisplatin 80 mg/sqm i.v. on day 1 and melphalan 12 mg/sqm i.v. on day 2 q 4 weeks. Response was verified by second-look laparotomy. WAR was carried out with the open field technique on a linear accelerator (daily dose: 1.3 Gy, total dose: 29.9 Gy) in patients with pathological or clinical CR or pathological PR with microscopical residual disease. RESULTS: 146/218 patients (67%, 95% CI: 61%-73%) responded to PAMP: 56 (26%) achieved pCR, 24 (11%), cCR, 56 (26%) pPR and 10 (5%) cPR (c = clinical, p = pathological). Multivariate analyses revealed that in advanced stages (92 cases in remission), the achievement of pCR was the most important factor for longer time to failure (TTF) and survival. Only 51/118 (43%) patients in remission received WAR. Early-stage patients <= 55 years were more likely to have WAR than patients older than 55 years (77% vs. 23%; p = 0.02). Advanced-stage patients with cCR were less likely to be irradiated than patients with pCR or pPR (10% vs. 51%; p = 0.003). Toxicity of PAMP was acceptable with 10% of WHO grade 4 hematologic toxicity. Acute hematological toxicity of WAR caused interruption (33%) or incompleteness (33%) of irradiation in the majority of patients. CONCLUSIONS: PAMP is an effective treatment for advanced ovarian cancer with a 67% response rate after 4 cycles. For the majority of patients in remission, WAR as a consolidation treatment was hardly feasible. For these patients new treatment modalities to consolidate remission are needed.