We conducted a double-blind treatment study of 110 adults from the United States who were attending summer classes in Guadalajara, Mexico, and had diarrhea (four or more unformed stools in 24 hours, or three or more unformed stools per eight-hour period plus one or more additional clinical indicators of enteric infection). Thirty-seven patients received trimethoprim/sulfamethoxazole (TMP/SMX) (160 mg of TMP and 800 mg of SMX), 38 were given TMP alone (200 mg), and 35 took a placebo twice daily for five days. By the end of the first 24 hours of treatment, patients taking either TMP/SMX or TMP alone passed fewer unformed stools than did patients given placebo (P = 0.0002 and P = 0.01, respectively). Abdominal pain and nausea were reduced in both treatment groups. The beneficial effect was seen in treatment of Escherichia coli-induced diarrhea, shigellosis, and diarrhea not associated with an enteropathogen. Five per cent of patients given TMP/SMX, 8 per cent of those given TMP, and 49 per cent of those given placebo were considered treatment failures (P less than 0.001 for both active drugs as compared with placebo). Early treatment with TMP/SMX or TMP is an alternative to prophylactic use of drugs for travelers' diarrhea.