The clinical usefulness of antibiotic susceptibility testing of fresh clinical isolates of anaerobes (primarily from blood cultures) was studied. Analysis of 65 patients showed that susceptibility reports were used in only 13 instances (20%), representing mainly orthopedic and CNS infections. Of the 47 patients were susceptibilities were not used, 20 received therapy (appropriate in each case) based on the culture report and 27 were treated empirically. Only 6 patients in the empirically treated group received inappropriate treatment, but 4 of those 6 died, and patients in this group as a whole had a worse outcome than did patients in the other groups. These empirically treated patients also had a somewhat worse prognosis. Susceptibilty testing of anaerobes may be reserved for bacteremic patients and for managing severe, chronic anaerobic infections, such as septic arthritis, osteomyelitis and brain abscesses.