Forty-five children (aged 1 day to 15 years) with bacterial meningitis were randomized to receive either traditional therapy (ampicillin and chloramphenicol or gentamicin, pending sensitivity) or ceftriaxone (100 mg/kg per day in two doses for a minimum of 10 days). The etiological agents involved were similar for the two groups and included Haemophilus influenzae type b, Neisseria meningitidis, Streptococcus pneumoniae, and group B streptococcus. Repeat spinal taps were carried out 24 to 48 h after admission. Organisms were seen on the Gram stain of one patient treated with ceftriaxone, but five patients in the traditional therapy group had organisms present on Gram stain of uncentrifuged spinal fluid or positive cultures of the spinal fluid (or both). Ceftriaxone entered the cerebrospinal fluid well, and the average cerebrospinal fluid bactericidal activity for ceftriaxone 1 h after a dose was at least 60 times greater than for ampicillin or chloramphenicol. In those patients who received treatment for a long enough period of time to permit evaluation, there was one death in each group, both due to S. pneumoniae. The length of fever and complications were similar for the patients in both groups. Ceftriaxone was well tolerated; diarrhea, seen in 5 of the 22 patients who received the drug, was the most commonly encountered adverse effect. It was mild, and in no case was it necessary to discontinue the drug. Ceftriaxone appears in this preliminary study to be a safe and acceptable single agent for the treatment of bacterial meningitis in children.