All 426 patients with ovarian malignancies registered in the population-based Tumor Registry of the Southeast region of Sweden during 1984 to 1987 were analyzed by survey of the hospital records and population registry data. After comparison with other population-based materials, it seems that the overall survival figures have improved in ovarian cancer. Excluding patients diagnosed at autopsy a 5-year corrected survival of 43% was recorded. Among patients aged under 45 years the corrected 5-year survival was 72%. In a Cox's regression analysis age and stage were significant predictors of cancer death while histology (epithelial vs. non-epithelial), although significant in the univariate analysis, did not add prognostic information in the multivariate model. The relative cancer death rate was 6.4 for patients aged over 74 years compared with those aged under 45 (p 1 cm relative to