Ceftriaxone was compared with cefotaxime for the treatment of serious bacterial infections in a prospective, randomized, double-blind clinical trial. The dose of ceftriaxone was 2 g once a day, and the dose of cefotaxime was 2 g every 4 h. Metronidazole was added if anaerobic infection was suspected. Explicit criteria were used to define infections, clinical response, and adverse effects. Ceftriaxone was given to 88 patients and cefotaxime to 83. The two treatment groups did not differ in types of infection, infecting organisms, and severity of underlying disease. The response rate was 81% (71/88) for ceftriaxone and 80% (66/83) for cefotaxime. The power of the study to detect a 15% difference in response rate at P less than .1 was 90%. The frequency of diarrhea, thrombophlebitis, prothrombin time, prolongation, colonization, and superinfection did not differ between treatment groups. Ceftriaxone 2 g once a day was as safe and effective as cefotaxime 2 g every 4 h for suspected serious bacterial infections.