The purpose of the experiments described here was to test whether membrane-impermeant antibiotics present in the extracellular milieu could kill bacteria within macrophages. For this, mouse macrophage hybrids and elicited mouse peritoneal macrophages first were allowed to phagocytose the facultative intracellular bacterium Listeria monocytogenes. The cells were incubated with or without gentamicin, and their bactericidal activity was measured. The results show that gentamicin caused normally nonbactericidal macrophages to kill L. monocytogenes. In addition, gentamicin caused listericidal cells to kill significantly more bacteria. To determine whether gentamicin accumulated within macrophages during culture, we tested whether lysates of macrophage hybrids cultured for 72 h in gentamicin-containing medium and then washed could kill Listeria cells. When cultured with 50 to 100 micrograms of gentamicin per ml, but not when cultured with 0 to 5 micrograms of gentamicin per ml, cell lysates were extremely listericidal, demonstrating the presence of intracellular gentamicin. Because gentamicin does not penetrate cell membranes, we hypothesized that it can be internalized by the cell through pinocytosis and can enter the same intracellular compartment as does phagocytosed L. monocytogenes. To test this, macrophages which had phagocytosed L. monocytogenes were incubated with the fluorochrome lucifer yellow to trace pinocytosed medium. About half of the Listeria cells within the macrophages were surrounded by lucifer yellow, indicating delivery of pinocytosed fluid, which could contain antibiotics, to phagosomes containing bacteria. The experiments described here indicate that membrane-impermeant antibiotics can enter macrophages and kill intracellular bacteria. Thus, the use of gentamicin in macrophage bactericidal assays can interfere with the results and interpretation of experiments designed to study macrophage bactericidal activity.