We compared the clinical and bacteriologic response of 5-day treatment with cefixime, 8 mg/kg per day, with the response to trimethoprim-sulfamethoxazole (TMP-SMX), 10-50 mg/kg per day, the currently recommended therapy. Of the assessable children with acute, culture-proven shigellosis, 38 received cefixime and 39 received TMP-SMX. Pretreatment data on the two study groups were similar. In the first group, all isolates were susceptible to cefixime; in the TMP-SMX group, 32 isolates were resistant and 7 were susceptible to TMP-SMX. Clinical response (day 5) showed cure, improvement, and failure in 89%, 8%, and 3%, respectively, of the cefixime group, and in 25%, 44%, and 31%, respectively, of the TMP-SMX-resistant group (p < 0.001). Bacteriologic cure (day 3) occurred in 78% and 23% of the cefixime and TMP-SMX-resistant groups, respectively (p < 0.001). Clinical or bacteriologic relapse (day 12) was infrequent in both groups. The response to treatment of the cefixime and the TMP-SMX-susceptible groups was similar. No significant side effects were noted. We conclude that cefixime is superior to TMP-SMX in the treatment of suspected shigellosis in areas with a high rate of resistance to TMP-SMX.