This prospective phase IV study on cohort concerns a vaccine made of the phenol-insoluble fraction of Brucella abortus biotype 1 (B19 strain). Three hundred and three professionally exposed subjects entered the study; 161 out of 182 subjects (88.5 percent) with negative response to an intradermal test for detection of previous contamination accepted to be vaccinated. Booster injections were given 18 and 36 months after vaccination. Local pain was observed after 45.2 percent of injections and moderate systemic reactions after 5 percent of injections. Seropositivity after primary vaccination reached 80 percent. The booster injection, justified by a major decrease of this rate after 18 months, gave exactly the same response of the thymo-independent type. This vaccinal schedule did not result in detectable hypersensitivity. The clinical effectiveness of the vaccine could not be evaluated accurately because of the insufficient number of subjects. The possibility of subclinical infection in vaccinated subjects calls for wider comparative studies of vaccinated versus non-vaccinated subjects.