We conducted a randomized trial among persons in rural Bangladesh to evaluate the side effects and immunogenicity of orally administered B subunit-killed whole cell (BS-WC) and killed whole cell-only (WC) cholera vaccines and a killed Escherichia coli strain K12 placebo proposed for field testing. Three doses of BS-WC, WC, E. coli, or a control agent were given with antacid to 1,257 women (aged greater than 15 years) and children (aged to to 15 years). The four groups exhibited no statistically significant differences in occurrence of symptoms after each dose, and rises in titers of vibriocidal (VC) antibodies to Inaba and Ogawa were twofold higher for vaccinees than for controls (P less than .001). Half of the persons with fourfold or greater VC responses to WC responded after the first dose; many additional patients, particularly young children, responded after subsequent doses. In contrast, 89% of persons who responded to BS-WC with twofold or greater rises in titer of IgG antibodies to cholera toxin did so after the first dose. After the third dose, vaccinees exhibited a fivefold higher rise in titer than did controls (P less than .001); a dose-to-dose booster effect was most evident in young children.