Information on 62 bacteriologically confirmed cases of bacterial meningitis treated with cefotaxime in this country was obtained retrospectively from infectious disease consultants. This series of cases differed markedly from the world cumulative case data thus far presented. One of the two most common organisms treated was the pneumococcus (allergy to penicillin or misdiagnosis of the Gram stain results were the major reasons given). The other organism was Klebsiella. Unanticipated bacteriologic successes were noted in two cases of staphylococcal meningitis secondary to parameningeal foci. The bacteriologic cure rate and survival rate were about 85 percent. Failure of monotherapy was seen in one case of Pseudomonas meningitis, as well as in three of five cases of Enterobacter meningitis. In addition, two cases of Escherichia coli meningitis in which moxalactam therapy inexplicably failed were cured with cefotaxime. Close analysis of killing kinetics appeared to explain the Enterobacter and E. coli failures. Thus, overall not all gram-negative species and not all isolates of any particular species that cause meningitis can be successfully treated by cephalosporins. Data obtained during the investigative trials do not appear to be entirely predicative of what occurred during the free clinical use of an antibiotic. Post-investigatory follow-up and surveillance of all newly introduced therapeutic agents are needed.