Taxol, a microtubule-binding diterpene, mimics many effects of lipopolysaccharide (LPS) on mouse macrophages. The LPS-mimetic effects of taxol appear to be under the same genetic control as responses to LPS itself. Thus we have postulated a role for microtubule-associated proteins (MAP) in the response of macrophages to LPS. Stimulation of macrophages by LPS quickly induces the activation of mitogen-activated protein kinases (MAPK). MAPK are generally considered cytosolic enzymes. Herein we report that much of the LPS-activatable pool of MAPK in primary mouse peritoneal macrophages is microtubule associated. By immunofluorescence, MAPK were localized to colchicine- and nocodazole-disruptible filaments. From both mouse brain and RAW 264.7 macrophages, MAPK could be coisolated with polymerized tubulin. Fractionation of primary macrophages into cytosol-, microfilament-, microtubule-, and intermediated filament-rich extracts revealed that approximately 10% of MAPK but none of MAPK kinase (MEK1A and MEK2) was microtubule bound. Exposure of macrophages to LPS did not change the proportion of MAPK bound to microtubules, but preferentially activated the microtubule-associated pool. These findings confirm the prediction that LPS activates a kinase bound to microtubules. Together with LPS-mimetic actions of taxol and the shared genetic control of responses to LPS and taxol, these results support the hypothesis that a major LPS-signaling pathway in mouse macrophages may involve activation of one or more microtubule-associated kinases.