A trial to evaluate the immune response to anti-meningococcal polysaccharides vaccines produced by the Instituto Vital Brazil was carried out in 1974. Two hundred and thirty-eight volunteers - employees and prisoners - of a penitentiary were allocated randomly to six groups, three of which received different types of the experimental vaccines (type A, C and A+C combined). A fourth group was inoculated with a placebo (tetanus toxoid). The last two groups were inoculated with type A or C vaccine already in use in the country to control the epidemic in the city of São Paulo. Blood samples were taken from each volunteer before and four weeks after the injection. A passive haemagglutination test was performed. The staff of the Instituto Adolfo Cruz responsible for the tests was not informed about the group origin of a sample or if it had been collected before or after the injection. The experimental vaccines containing polysaccharide type A (monovalent and combined with type C) did not show specific antigenic activity. In fact, a paradoxical formation of anti-C antibodies after the inoculation of monovalent type A vaccine was observed. On the other hand the two groups inoculated proportion of collateral effects, although very mild, reflecting technical problemas of productions. The type C vaccine under trial, however, gave excellent results. Besides the main objective of the trial, the investigation gave rise to interesting observations about polysaccharide vaccines in general. Firstly the C type vaccines gave higher conversion rates and differencial titres than the A type ones. Secondly there was a fali in the level of non-specific antibodies tested (against A if the vaccine injected was type C) strengthening the by other studies, which deserves further studying.