A live oral attenuated vaccine, lacking the enzyme epimerase, has been given with complete safety to 173 men. Two preparations of the vaccine, one with and one without galactose in the initial culture media, led to somewhat differing results. Vaccine A, prepared with galactose, was more readily identified in the stool specimens of its recipients and was more likely to provoke an O antibody response than was vaccine B. Furthermore, the clinical protection from typhoid fever was significant in the vaccine A group, whereas not so with the vaccine B counterpart. The protection afforded by vaccine A exceeds that of any of our previous candidate strains. It is perhaps not presumptuous to imagine that such a vaccine would be equally effective in an area endemic for typhoid fever, where the vaccine might act as a booster effect in a previously exposed population. Remaining questions to be answered include the duration of the protection, and the efficacy of a lyophilized preparation.