The pathogenesis of ovarian carcinoma, the most lethal gynecological malignancy, is unknown because of the lack of a tumor progression model. Based on a review of recent clinicopathological and molecular studies, we propose a model for their development. In this model, surface epithelial tumors are divided into two broad categories designated type I and type II tumors that correspond to two main pathways of tumorigenesis. Type I tumors tend to be low-grade neoplasms that arise in a stepwise manner from borderline tumors whereas type II tumors are high-grade neoplasms for which morphologically recognizable precursor lesions have not been identified, so-called de novo development. As serous tumors are the most common surface epithelial tumors, low-grade serous carcinoma is the prototypic type I tumor and high-grade serous carcinoma is the prototypic type II tumor. In addition to low-grade serous carcinomas, type I tumors are composed of mucinous carcinomas, endometrioid carcinomas, malignant Brenner tumors, and clear cell carcinomas. Type I tumors are associated with distinct molecular changes that are rarely found in type II tumors, such as BRAF and KRAS mutations for serous tumors, KRAS mutations for mucinous tumors, and beta-catenin and PTEN mutations and microsatellite instability for endometrioid tumors. Type II tumors include high-grade serous carcinoma, malignant mixed mesodermal tumors (carcinosarcoma), and undifferentiated carcinoma. There are very limited data on the molecular alterations associated with type II tumors except frequent p53 mutations in high-grade serous carcinomas and malignant mixed mesodermal tumors (carcinosarcomas). This model of carcinogenesis reconciles the relationship of borderline tumors to invasive carcinoma and provides a morphological and molecular framework for studies aimed at elucidating the pathogenesis of ovarian cancer.