We performed case-control analyses using data from the North Carolina Ovarian Cancer Study to determine risk factors that distinguish primary peritoneal cancer (PPC) from epithelial ovarian cancer (EOC). Our risk factor analyses were restricted to invasive serous cancers including 495 EOC cases, 62 PPC cases and 1,086 control women. Logistic regression analyses were used to calculate adjusted odds ratios and 95% confidence intervals for risk factor associations. Although many case-control associations for the invasive serous PPC cases were similar to those of the invasive serous EOC cases, some differences were observed including a twofold increase in risk of invasive serous PPC in women who were >or=35 years at last pregnancy, whereas a decreased risk was observed for invasive serous EOC risk. We could not confirm a previous report of an association between tubal ligation and PPC, a factor consistently associated with a decreased risk of EOC. The difference in the risk factor associations between invasive serous PPC and EOC cancers suggests divergent molecular development of peritoneal and ovarian cancers. A larger study to determine risk factors for invasive serous PPC is warranted.