We previously presented evidence [Nagy et al. (1997) Neuroscience 78, 533-548] that, in addition to their ubiquitous expression of connexin43, astrocytes produce a second connexin suggested to be connexin30, a recently discovered member of the family of gap junction proteins. A connexin30 specific antibody was subsequently developed and utilized here to confirm and extend our earlier observations. On western blots, this antibody detected a 30,000 mol. wt protein in rat, mouse, cat and human brain, and exhibited no cross-reaction with connexin43, connexin26 or any other known connexins expressed in brain. Immunohistochemically, connexin30 was localized in astrocytes, at gap junctions between these cells and on the astrocyte side of gap junctions between astrocytes and oligodendrocytes. Double labelling revealed co-localization of connexin30 and connexin43 at astrocytic gap junctions. Punctate immunolabelling patterns for both connexins were qualitatively similar, but differences were evident. In contrast to regional connexin43 expression, diencephalic and hindbrain areas exhibited considerably greater expression than forebrain areas, subcortical perivascular astrocytic endfeet were more heavily labelled for connexin30, white matter tracts such as corpus callosum, internal capsule and anterior commissure were devoid of connexin30, and appreciable levels of connexin30 during development were not seen until about postnatal day 15. These results indicate that connexin30 is expressed by gray, but not white matter astrocytes, its distribution is highly heterogeneous in gray matter, it is co-localized with connexin43 at astrocytic gap junctions where it forms homotypic or heterotypic junctions, and its emergence is delayed until relatively late during brain maturation. Taken together, these results suggest that astrocytic connexin30 expression at both regional and cellular levels is subject to regulation in adult brain as well as during brain development.