During the months of July 1977 and July 1985, students from the United States participated in a double-blind, placebo-controlled trial examining the effectiveness of liquid bismuth subsalicylate (BSS) (1977) and two dosages of the tablet formulation of BSS (1985) in preventing diarrhea while in Guadalajara, Mexico. In the first study, 62 subjects received BSS for 3 weeks at a dosage of 60 mL four times daily (4.2 g of BSS/d) compared with 66 students receiving an oral placebo at a similar dosage schedule. In the second study, 51 students took two tablets four times daily (2.1 g of BSS/d), 63 took one tablet four times daily (1.05 g of BSS/d), and 58 took a placebo (two tablets taken four times daily), each for 3 weeks. In the initial study, 14 (23%) BSS-treated subjects developed diarrhea compared with 40 (61%) placebo-tested persons (P less than .0001). In the second trial, seven (14%) subjects taking two tablets of BSS four times daily, 15 (24%) taking one tablet of BSS four times daily, and 23 (40%) receiving placebo tablets experienced diarrhea (P less than .001 for the higher dose). The percent protection provided by BSS was 62% for the group that received 4.2 g/d, 65% for 2.1 g/d, and 40% for 1.05 g/d, when compared with the corresponding placebo group. In cases in which stools were analyzed, seven (24%) of 29 BSS-treated subjects who had diarrhea had a detectable enteric pathogen, compared with 35 (59%) of 59 of those randomized to receive a placebo. BSS was well tolerated.(ABSTRACT TRUNCATED AT 250 WORDS)