To compare the therapeutic efficacy of ceftriaxone given once daily for 5 days and chloramphenicol given four times daily for 14 days, a controlled trial was carried out with 59 patients who were culture positive for Salmonella typhi. Ceftriaxone was given to 28 patients in once-daily intravenous doses of 75 mg/kg of body weight to children and 4 g to adults for 5 days; chloramphenicol was given to 31 patients at a dosage of 60 mg/kg/day until defervescence and then at 40 mg/kg/day to complete 14 days of treatment. All Salmonella isolates were susceptible to both antibiotics. Clinical cures (defervescence without complications, no relapse, and no need for further treatment) occurred in 79% of the patients treated with ceftriaxone and 90% of those treated with chloramphenicol (P = 0.37). On the third day of treatment, blood cultures were positive for S. typhi for 60% of the patients in the chloramphenicol group and 0% of the ceftriaxone group (P = 0.001). Defervescence occurred in half the patients in both groups during the first 7 days, but on days 9 to 13 after the start of treatment, nine patients in the ceftriaxone group, compared with six patients in the chloramphenicol group, remained febrile (P = 0.4). The median hematocrit and total leukocyte counts at day 14 were significantly lower for the chloramphenicol group than those for the ceftriaxone group (P = 0.01 and P = 0.02, respectively). These results indicate that the effects of therapy with ceftriaxone for typhoid fever differed from those of chloramphenicol therapy in that blood cultures became negative earlier, prolonged fever persisted in some patients, and bone marrow suppression was reduced. We conclude that a short, 5-day course of ceftriaxone is a useful alternative to conventional 14-day chloramphenicol therapy in the treatment of typhoid fever.