The vole agent described by Baker in 1946 was studied as an example of a bacterium that has been mistakenly regarded a rickettsia. Unlike rickettsiae, the vole agent killed chicken embryos with great irregularity, multipled primarily at the surface of avian or mammalian cells and not intracellularly, produced colonies rather than plaques on chicken embryo monolayers under agar, and developed small colonies after 4 to 7 days of cultivation on blood plates. It was most conveniently cultivated on monolayers of irradiated L cells and was purified by minor modifications of the Renografin gradient procedure used for rickettsiae. It actively catabolized glutamate, glutamine, succinate, and pyruvate, but not glucose or glucose-6-phosphate. Enzymatic activities of cell extracts were consistent with above findings. The base ratio (molar percent guanine plus cytosine) of its deoxyribonucleic acid was shown to be 39, which was identical to the base ratio of the deoxyribonucleic acid of Rochalimaea quintana tested simultaneously. Serological studies indicated no cross-reactivity with Rickettsia tsutsugamushi, but strong cross-reaction with R. quintana was observed when a hyperimmune rabbit serum and a convalescent human serum were tested. We conclude that the vole agent is a strain of the trench fever rickettsia, R. quintana.