We report a case of Mycobacterium xenopi infection of the spine in a 70-year-old woman. The findings of our case and of five other published reports of bone or joint infection with M. xenopi illustrate the problems with diagnosis and management of infection with this organism. Detection of M. xenopi in clinical specimens may require prolonged incubation at 37 degrees C or incubation at higher temperatures. Furthermore, M. xenopi may be misidentified as Mycobacterium avium-Mycobacterium intracellulare complex. For these reasons, human infection due to M. xenopi may be underrecognized. The optimal therapy for human infection with M. xenopi is unknown.