The etiology and early events in the progression of epithelial ovarian carcinomas are among the least understood of all major human malignancies. There are no adequate means for early detection of these neoplasms and, as a result, they are usually diagnosed in late stages. The purpose of this review is to point out some of the peculiar problems and limitations that have hampered progress in ovarian carcinogenesis research and to summarize new approaches and recent advances in our understanding of this process. The review first presents an overview of the properties of the ovarian surface epithelium (OSE) which is thought to be the source of epithelial ovarian carcinomas, followed by a discussion of recent research based on human OSE. This includes sections on methodology for the attainment and study of OSE, investigations of OSE from women with predisposing mutations, and attempts to convert normal OSE to malignancy. This overview is followed by a discussion of the contributions, potential, and limitations of animal models. The knowledge gained by these approaches will likely lead to improvements in our ability to prevent, diagnose, and treat ovarian cancer.