Anaerobic bacteria commonly cause soft tissue infections in humans, usually as a result of trauma. Although the susceptibility of some species to antibiotics can be predicted reasonably well, many clinical laboratories can do susceptibility tests more easily than they can correctly identify anaerobes. The broth-disk test is the simplest test for clinical laboratories, and a new standardized MIC method is now available for research laboratories. Several types of experimental anaerobic infections can be initiated in animals, and these infections have been used to test the efficacy of therapeutic regimens. Metronidazole was the most effective, and clindamycin was also very active. Elimination of the anaerobic component of abscesses may be sufficient therapy; it may be possible to ignore the facultative organisms.