MAP kinase (mitogen activated protein kinase) represents a ubiquitously expressed family of kinases whose long term activation via phosphorylation is essential for the mitogenic response in fibroblasts. Two family members, p42 and p44 MAP kinase are cytosolic proteins in quiescent cells, but become nuclear following mitogenic stimulation. Inactivation of MAP kinases occurs via a specific phosphatase, MKP-1. Hence, we examined the localisation of this phosphatase, to determine the cellular site of MAP kinase inactivation. Transient transfection of CCL39 fibroblasts with epitope-tagged MKP-1 showed the protein to be entirely nuclear in both quiescent and mitogen stimulated cells, whereas a catalytically inactive mutant in which the essential cysteine was mutated to serine (MKP-1CS) was predominately cytoplasmic and again serum stimulation failed to alter the protein's localisation. Expression of either wild type or inactive MKP-1 did not alter the cytosolic localisation of p44 MAP kinase in quiescent cells nor the ability of MAP kinase to translocate to the nucleus following mitogen stimulation. Expression of wild type MKP-1 inhibited serum stimulated early (c-fos promoter) and late (dhfr promoter) transcriptional events as well as entry into S-phase. This inhibition was reversed by the co-expression of an active MAP kinase. We conclude that in the continual expression of MKP-1, the cellular localisation of MAP kinase is unaffected and that inactivation of MAP kinase by MKP-1 is a nuclear process leading to the inhibition of cell division.