A compilation of literature data and recent experiments led to the following conclusions regarding cyclic adenosine 3':5' monophosphate (cAMP) regulation of gene expression. Several classes of cAMP-induced gene expression can be discriminated by sensitivity to stimulation kinetics. The aggregation-related genes respond only to nanomolar cAMP pulses. The prestalk-related genes respond both to nanomolar pulses and persistent micromolar stimulation. The prespore specific genes respond only to persistent micromolar stimulation. The induction of the aggregation- and prestalk-related genes by nanomolar cAMP pulses may share a common transduction pathway, which does not involve cAMP, while involvement of the inositol 1,4,5-trisphosphate (IP3)/Ca2+ pathway is unlikely. Induction of the expression of prespore and prestalk-related genes by micromolar cAMP stimuli utilizes divergent signal processing mechanisms. cAMP-induced prespore gene expression does not involve cAMP and probably also not cyclic guanosine 3'.5' monophosphate (cGMP) as intracellular intermediate. Involvement of cAMP-induced phospholipase C (PLC) activation in this pathway is suggested by the observation that IP3 and 1,2-diacylglycerol (DAG) can induce prespore gene expression, albeit in a somewhat indirect manner and by the observation that Li+ and Ca2+ antagonists inhibit prespore gene expression. Cyclic AMP induction of prestalk-related gene expression is inhibited by IP3 and DAG and promoted by Li+, and is relatively insensitive to Ca2+ antagonists, which indicates that PLC activation does not mediate prestalk-related gene expression. Neither prespore nor prestalk-related gene expression utilizes the sustained cAMP-induced pHi increase as intracellular intermediate.