We used co-cultures of porcine ovarian granulosa cells and mouse adrenocortical tumor cells (Y-1) to examine the kinetics of contact-dependent intercellular signal transfer and to assess the molecular mechanisms employed by this process. Exposure to follicle-stimulating hormone (FSH) caused cAMP-dependent protein kinase dissociation in granulosa cells and, with time, in Y-1 cells if, and only if, they contacted a responding granulosa cell. Y-1 cells close to a granulosa cell but not touching it failed to respond similarly. In reciprocal experiments, co-cultures were stimulated with adrenocorticotropic hormone (ACTH). Y-1 cells dissociated protein kinase as did granulosa cells in contact with Y-1 cells; however, granulosa cells that were not in contact with Y-1 cells failed to respond to the hormone. Fluorogenic steroids were secreted by Y-1 cells cultured alone and stimulated with ACTH, but were not secreted by cultures exposed to FSH. Neither hormone caused fluorogenic steroid production by granulosa cells. On the other hand these steroids were secreted in co-cultures stimulated with ACTH and to a lesser degree in co-cultures exposed to FSH. Autoradiography revealed that I125-FSH bound only to granulosa cells, never to Y-1 cells, even if they were in contact with an ovarian cell. The possibility of cell fusion was tested by experiments in which Y-1 cell membranes were labeled with cationized ferritin. These cells were then placed in co-culture with ovarian granulosa cells that had previously been allowed to ingest latex spheres. At regions of gap junctions between Y-1 and granulosa cells ferritin remained attached to the adrenal cell membrane and was never observed to migrate to the granulosa cell membrane. From these data, we conclude that hormone specific stimulation of one cell type leads to protein kinase dissociation in heterotypic partners only if they contact a hormone responsive cell. This signal transfer is bidirectional, exhibits temporal kinetics and occurs in the absence of apparent cell fusion. The only structural feature connecting Y-1 and granulosa cells were gap junctions implying they provided the communication channels; however, alternative mechanisms cannot be excluded. We have not established the identity of the signal being transferred although cAMP is a logical candidate.