In Reply—The cases of meningitis that we presented do not prove a cause-and-effect relationship between lumbar punctures (LPs) and meningitis. As was stated clearly in the article, the "most obvious possibility is that the patient was bacteremic and had seeded the meninges prior to the initial examination." Since an LP must be performed on every child whose clinical condition suggests sepsis or meningitis, a prospective, controlled study would be unethical and immoral. Carefully controlled studies in animals, however, have demonstrated that, in general, experimental bacteremia alone does not produce meningitis.1-5 When the animals were subjected to lumbar puncture during bacteremia, meningitis was regularly observed.4,6,7 It is clear that the vast majority of cases of meningitis are in no way associated with a previous LP. The combined clinical and experimental data would suggest that under certain circumstances an initial LP might enhance the development of meningitis.