We examined 92 patients with epithelial ovarian cancer and 262 patients with benign ovarian diseases undergoing laparotomy. On the basis of a nonparametric method, antigen levels corresponding to prefixed 95% specificity values in a group of 674 women with benign gynecologic diseases were taken as cutoff limits (88.8 U/ml for CA 125 and 13.7 U/ml for CAM 29). Moreover, CA 125 and CAM 29 levels were measured serially during and after chemotherapy in 26 women selected from the patients with advanced epithelial ovarian cancer. At diagnosis, serum CA 125 was as sensitive as serum CAM 29 for nonmucinous tumors, but more sensitive than serum CAM 29 for mucinous tumors. The association of the two markers seemed to give no advantage over the CA 125 assay alone in the diagnosis of epithelial ovarian cancer. In monitoring the response to chemotherapy and follow-up of patients with epithelial ovarian cancer, changes in CA 125 levels correlated with the clinical course of disease better than changes in CAM 29 levels, and the serum CA 125 assay was more reliable than the serum CAM 29 assay in the early detection of tumor progression. In conclusion, serum CAM 29 did not seem to represent a complementary assay to serum CA 125 in the management of patients with epithelial ovarian cancer.