The ability of three analogs of ACTH1-24 ([Gln5, Phe9] ACTH1-24, [Gln5, Ala9[Acth1-24, and [Gln5, Lys8, Phe9[ ACTH1-24) embodying tryptophan substitutions to activate the adenylate cyclase system of a bovine adrenal plasma membrane preparation was compared to the effect of the analogs on adenosine 3':5'-monophosphate (cyclic AMP) accumulation and steroidogenesis in viable bovine adrenocortical cells. The results were not comparable. Whereas the analogs antagonized the ACTH1-24-activated membrane cyclase they stimulated cyclic AMP accumulation as well as steroid production of the cells. None of the analogs inhibited steroidogenesis of ACTH1-24-stimulated cells, but two of them, at very high dose levels, inhibited cyclic AMP production. The ability of the analogs to stimulate steroidogenesis of the adrenal cells half-maximally decreased in the order tryptophan greater than phenylalanine greater than alanine, indicating that the aromaticity of the indole ring of tryptophan is necessary for maximal interaction between hormone and receptor. Both the absolute and relative steroidogenic potencies were the same for several analogs when assayed with rat adrenal cells. Although only a small fraction of the cell's potential to produce cyclic AMP was necessary to induce maximum steroid production, the relative activities of a series of analogs were the same for steroidogenesis as for cyclic AMP accumulation. Furthermore, the concentration of cyclic AMP necessary for full steroidogenesis was practically identical for a series of peptides that differed widely in potency. These findings support the postulate that cyclic AMP accumulation and steroidogenesis in adrenocortical cells are coupled processes. The differential behavior of bovine adrenal plasma membranes and bovine adrenocortical cells toward ACTH analogs indicates that structure-function studies using cyclase assays may not reflect events that take place in the intact adrenal or in cell preparations derived therefrom.