The regulated dephosphorylation of mitogen-activated protein kinases (MAPKs) plays a key role in determining the magnitude and duration of kinase activation and hence the physiological outcome of signalling. In mammalian cells, an important component of this control is mediated by the differential expression and activities of a family of 10 dual-specificity (Thr/Tyr) MAPK phosphatases (MKPs). These enzymes share a common structure in which MAPK substrate recognition is determined by sequences within an amino-terminal non-catalytic domain whereas MAPK binding often leads to a conformational change within the C-terminal catalytic domain resulting in increased enzyme activity. MKPs can either recognize and inactivate a single class of MAP kinase, as in the specific inactivation of extracellular signal regulated kinase (ERK) by the cytoplasmic phosphatase DUSP6/MKP-3 or can regulate more than one MAPK pathway as illustrated by the ability of DUSP1/MKP-1 to dephosphorylate ERK, c-Jun amino-terminal kinase and p38 in the cell nucleus. These properties, coupled with transcriptional regulation of MKP expression in response to stimuli that activate MAPK signalling, suggest a complex negative regulatory network in which individual MAPK activities can be subject to negative feedback control, but also raise the possibility that signalling through multiple MAPK pathways may be integrated at the level of regulation by MKPs.