Cyclic AMP as well as the specific activity of cyclic AMP-dependent protein kinase decreased from the first two hours after Chinese hamster ovary cells in plateau phase were stimulated to proliferate by tripsinization of confluent cultures and dilution in fresh media. From two to five hours after this stimulation, the cyclic AMP level and the specific activity of cyclic AMP-dependent protein kinase increased two-fold. There was a 40--50% increase in the degree of activation of cyclic AMP-dependent protein kinase during this same time interval. In plateau cultures prior to being stimulated to proliferate, type I cyclic AMP-dependent protein kinase was the predominant soluble form of these enzymes. At five hours after release from plateau, the predominant type of cyclic AMP-dependent protein kinase was type II. However, there was also a significant amount of type I present at this time. Types I and II cyclic AMP-dependent protein kinases were differentially detectable during the cell cycle of Chinese hamster ovary cells synchronized by mechanical selection of metaphase cells following colcemid treatment. During mitosis, type I kinase was predominant with only a small amount of type II activity detectable. The amount of activity of type I then progressively decreased as cells entered G1. During early G1, there was no detectable activity of type II kinase, but its activity increased from mid to late G1 and then decreased during the S phase. These data show a tight temporal relationship between the levels of cyclic AMP, the total cellular pool of type I and II cyclic AMP-dependent protein kinases, and the degree of activation of these kinases as cells traversed G1 toward S phase. These data suggest that the expression of each type of kinase may be important for the regulation of substrate phosphorylation during the cell cycle.