In the yeast Saccharomyces cerevisiae several phenotypic properties controlled by cAMP-dependent protein kinase (cAPK) are indicative of high cAPK activity during growth on glucose and low activity during growth on non-fermentable carbon sources and in stationary phase. It has been a matter of debate whether the apparently higher activity of cAPK in cells growing on glucose is due to a higher cAMP level or to an alternative mechanism activating cAPK. The cAMP level during diauxic growth of yeast cells in cultures with different initial glucose levels and different initial cell densities has been reinvestigated and the previously reported twofold increase in cAMP during growth initiation has been confirmed. However, this increase was transient and entirely associated with the lag phase of growth. The initiation of exponential growth on glucose was associated with a decrease in the cAMP level and there was no correlation between this decrease in cAMP and the depletion of glucose in the medium. In mutants defective in feedback inhibition of cAMP synthesis, resuspension of exponential-phase glucose-grown cells in glucose medium caused an extended lag phase during which a huge, transient accumulation of cAMP occurred. The latter required the presence of glucose and nitrogen, but not phosphate or sulfate, and was not due to intracellular acidification, as shown by in vivo 31P-NMR spectroscopy. The initiation of exponential growth on glucose was also associated in this case with a decrease in cAMP rather than an increase. This behaviour was also observed in strains with attenuated catalytic subunit activity and lacking the regulatory subunit and even in strains without catalytic subunits of cAPK. This might indicate that other mechanisms are able to cause down-regulation of cAMP synthesis in a way mimicking feedback inhibition. Transfer of glucose-growing cells of wild-type or cAPK-attenuated strains to a nitrogen starvation medium resulted in an increase in the cAMP level rather than a decrease. The results indicate that the apparent changes in cAPK activity in vivo during diauxic growth on glucose and during nitrogen starvation cannot be explained on the basis of changes in the cAMP level.