Vaccination is at present the only means of influenza control; so far, large-scale trials of live vaccine have been made mainly in the USSR. This paper discusses such a trial in persons above 12 years of age.About 40% of the population of Smolensk and about 50% of the population of the nearby town of Jarcevo were vaccinated with live influenza vaccine in the winters of 1964-65 and 1965-66, and the incidence of influenza and other acute respiratory diseases in these towns during the 1965 epidemic and the 1966 pre-epidemic period was compared with that in nearby "control" towns. Most subjects were vaccinated 2 or 3 times with divalent A2-B vaccine in 1964, but some only once; in 1965-66, most subjects were vaccinated once with monovalent B vaccine and once with divalent A2-B vaccine.Analysis of the incidence data for the towns involved, of more detailed incidence data for about 30 000 workers and 4000 schoolchildren in Smolensk and one control town, and of a controlled trial involving about 4000 persons, indicated that the large-scale vaccination led to a reduction in incidence of about 1.5- to 2-fold in 1965 and of about 2- to 3-fold in 1966. Limited serological studies in 1966 indicated that the reduction in incidence in the group studied was not 3-fold but 4-fold. The rather low protection offered by the 1964 vaccination may have been due to the low immunogenicity of the vaccine, or to the fact that the vaccine strains used did not correspond exactly to the influenza virus strains circulating in nature.