The catalytic (C) subunit of cAMP-dependent protein kinase is inhibited by the regulatory (R) subunit and by a thermostable inhibitor (PKI). Both inhibitors also affect the intracellular distribution of the C subunit. Whether injected into the cytoplasm or into the nucleus, free C subunit can enter and exit the nucleus freely. After 30 min its distribution is identical and is independent of the initial site of injection. In contrast, when C is injected into the cytoplasm complexed with R or PKI, the complexes are restricted to the cytoplasm (1-3). However, unlike the R subunit, which is restricted to the cytoplasm like the holoenzyme, free PKI enters the nucleus rapidly following its injection into the cytoplasm. When holoenzyme is injected directly into the nucleus, it cannot exit and return to the cytoplasm. In contrast, nuclear injection of a C.PKI complex results in the rapid exit of the C subunit from the nucleus. In equilibrated cells previously injected with the C subunit, subsequent cytoplasmic injection of either PKI or type 1 R depletes the nucleus of C although PKI does so faster, consistent with its ability to enter the nucleus. Both inhibitors block the cAMP response element-regulated gene expression. Hence PKI may serve as a nuclear scavenger of C providing a mechanism not only for inhibition but also for subcellular localization in the presence of cAMP by restricting the access of the C subunit to the nucleus.