Although ovarian cancer is the most common gynecological malignancy with a relatively poor 5-yr survival record, the mechanism(s) by which these tumors arise is not well understood. A role for inhibins and activins in regulating this transformation is suggested by the detection of circulating alpha or dimeric inhibin in some patients with ovarian cancer and by the alpha inhibin knockout mouse, in which development of gonadal tumors in 100% of homozygotes is associated with greatly elevated activin levels. To develop diagnostic tools with greater specificity for ovarian cancers, the present study was targeted at characterizing the biosynthetic capacity of the epithelial ovarian cancer cell lines from the American Type Culture Collection with respect to inhibin, activin, the related activin-binding protein follistatin (FS), and activin receptor type II. In addition, the functional capacity of this system was investigated by examining the ability of activin and FS to modulate cellular proliferation. All six cell lines contained abundant messenger RNA (mRNA) for activin receptor type II, but no inhibin alpha-subunit mRNA was detected in any cell line. Two cell lines contained mRNA for activin beta B-subunit (CaOV4 and SKOV3), one cell line contained beta A-subunit mRNA (SW626), and one cell line contained both (ES2); the latter also contained FS mRNA. FS mRNA was detected in another cell line (PA-1) that contained no detectable activin beta-subunit mRNA. Finally, one cell line (CaOV3) contained neither beta-subunit nor FS mRNA. Protein secretion was also examined. Consistent with the mRNA studies, the two cell lines containing FS mRNA secreted FS (PA-1 and ES2 cells), whereas three of the remaining lines secreted activin (A or B). In the cell line containing neither FS nor beta-subunit mRNA, no FS or activin could be detected. Finally, none of the cell lines secreted detectable immunoreactive inhibin. The effects of exogenous activin and FS on cellular proliferation were examined in these cell lines. No response was detected in the two cell lines that secreted FS (PA-1 and ES2). For the four cell lines not synthesizing FS, treatment with activin (1-100 ng/ml) resulted in an increase, whereas FS treatment (1-100 ng/ml) resulted in a decrease in cellular proliferation, as determined by [3H]thymidine incorporation. The response to activin correlated negatively with endogenous activin production, suggesting that autocrine activin production may be involved with cell proliferation. The differential expression and production of inhibin/activin subunits, activin receptors, and follistatin as well as the range of responses to exogenous activin among six ovarian epithelial cancer cell lines suggest that this family of hormones may be important in regulating cell proliferation in the ovary. Whether primary tumors have the same profile and the degree to which these results can be generalized to additional forms of ovarian cancer remain to be determined.