When determined under the usual conditions of an excess of ligand over protein, the concentration of cyclic AMP necessary to activate pure preparations of cyclic AMP-dependent protein kinase (EC 2.7.1.37; ATP:-protein protein phosphotransferase) half-maximally is in the range of 0.2-0.3 muM when casein or glycogen synthetase is used as the substrate, i.e., essentially the same as the concentration of the nucleotide that is found in resting skeletal muscle. The apparent dissociation constant for cyclic AMP bound to the protein kinase is also about 0.2-0.3 muM when measured under similar conditions. The concentration of the protein kinase in muscle is relatively high (0.23 muM), however, and under these conditions the apparent activation constant of the enzyme for cyclic AMP is raised so that an increase in cyclic AMP levels in the tissue would cause a concomitant increase in protein kinase activity over a wide range of nucleotide concentration. As a result, it is unnecessary to invoke compartmentalization of cyclic AMP to explain how it can control protein kinase activity in vivo. Another factor that may increase the effectiveness of changes in cyclic AMP concentration is the heat-stable protein inhibitor of protein kinase that may function to inhibit the activity of nearly all the protein kinase catalytic subunit dissociated by basal concentrations of cyclic AMP. Finally, the near equity between the concentration of cyclic AMP binding sites and the ligand itself provides a potential mechanism whereby agents can affect the total cyclic AMP content without directly affecting adenylate cyclase, cyclic AMP phosphodiesterase, or cyclic AMP transport.