The stimulation of incorporation of [3H]galactose into membrane glycoconjugates, measured in a precipitation test, was used as a criterion for activation of bone marrow cells. In this assay, purified bacterial lipopolysaccharide, lipoprotein, and murein monomer and dimer fragments all activated rat bone marrow cells in vitro. The response was dose dependent, followed a defined time course, and was not serum dependent. O-Acetylated murein dimer fragments from Proteus mirabilis were much less active than their unsubstituted counterparts, indicating a structural specificity for murein activation. Removal of adherent and phagocytizing cells from the marrow suspensions did not alter these results. The labeled, activated cells constituted a distinct population of buoyant density 1.064 to 1.069 g/cm3 when centrifuged on a continuous gradient of Percoll. Enrichment of the target cell population was achieved by a combination of adherent cell removal and discontinuous density gradient centrifugation to remove granulocytes and erythropoietic cells. It was concluded that a population of myelopoietic precursors could be activated by direct contact with bacterial cell wall constituents. The stimulation of galactose incorporation was not coupled to active deoxyribonucleic acid synthesis in the marrow cells. Thus, the activation was interpreted as an induction of differentiation rather than a mitotic event.