Activin A is produced by peritoneal mesothelial cells (PMCs), endometrial epithelial cells (EECs), and endometrial stromal cells (ESCs). Activin A levels are increased in the fluid of endometriotic cysts and activin receptors have been demonstrated in PMCs, eutopic endometrium, and endometriotic lesions. Collectively, these findings suggest a role for activin A in the pathogenesis of endometriosis. Here, we evaluated the effect of activin A on EEC and ESC invasion into the peritoneum using a novel in vitro model. In vitro study. Modeled peritoneum was established by growing the PMC line, LP9, to confluence over membranes with 8 micron holes that were coated with the solubilized growth factor reduced basement membrane preparation, Matrigel. EECs (n=8) and ESCs (n=8) were cultured as previously described. EECs or ESCs were treated for 24 hours with Activin A (6.25-100 ng/ml). EECs and ESCs were labeled with CellTracker® Green and placed on the PMC monolayer for 24 hours. Cells attached to the membrane surface (i.e. cells that did not invade), were mechanically removed and the membranes were placed in formaldehyde. Invaded cells on the bottom of the membrane were counted using a fluorescence microscope. In subsequent experiments, EECs and ESCs were treated for 24 hours with activin A (25 ng/ml) in combination with follistatin (250 ng/ml) or inhibin A (50 ng/ml), placed on the modeled peritoneum, and invasion was assessed at 24 hours. Attachment of ESCs and EECs to a PMC monolayer in the presence and absence activin A, activin A plus follistatin, and activin A plus inhibin A was compared using an established EEC and ESC/PMC attachment assay. The effect of activin A, activin A plus follistatin, and activin A plus inhibin A on ESC and EEC proliferation was assessed by measuring the degree of reduction of the yellow tetrazolium MTT salt producing intracellular purple formazan that was then solubilized and quantified spectrophotometrically. Activin A increased invasion through the modeled peritoneum in a dose dependent fashion. The 25 ng/ml dose resulted in a 2-fold (p<.05 vs control) and >2 fold (p<.04 vs control) increased EEC and ESC invasion, respectively. Follistatin and inhibin A blocked the activin A (25 ng/ml)-induced increase in invasion of both EECs and ESCs. There were no differences in the rate of attachment of EECs and ESCs to PMC monolayers with the activin A, activin A plus follistatin, or activin A plus inhibin A treatments. Likewise, there were no significant differences in the rate of proliferation of EECs or ESCs with activin A, activin A plus follistatin, or activin A plus inhibin A treatments. Activin A increases EEC and ESC invasion into modeled peritoneum. This effect of activin A on invasion was blocked by the addition of follistatin and inhibin A. The increased invasion following activin A treatment was not related to an increase in EEC- and ESC-PMC attachment or an increase in EEC/ESC proliferation. These findings suggest that activin A has an autocrine (EEC and ESC) and/or paracrine (through PMC binding of EEC and ESCs) effect that promote endometrial cell invasion into the submesothelial peritoneal extracellular matrix. Further investigations are warranted to characterize the roles of activin A and activin receptors in the development of the early endometriotic lesion.