To investigate the protective effects of pneumococcal vaccine, we assayed serum from healthy adults and from elderly bronchitics for antibody and opsonic activity against nine serotypes of S. pneumoniae. Before vaccination, there was no relation between opsonization and the level of antibody measured by RIA. Some serotypes were well opsonized in the absence of detectable antibody to capsular polysaccharide; others were not, despite modest levels of antibody. These in vitro studies did not support the concept that a certain level of antibody (e.g., >250 ng of antibody nitrogen/ml) was specifically associated with the capacity to opsonize pneumococci. Nearly all post vaccination sera had increased antibody and opsonic activity against all serotypes, but the lack of correlation in any individual serum persisted. RIA showed that pre- and postvaccination levels of antibody in elderly adults with chronic lung disease were similar to those of younger adults. In elderly bronchitics, opsonizing activity for six of the nine serotypes was lower after vaccination, a result suggesting a possible explanation for the failure of pneumococcal vaccine to be fully protective in these subjects. Elderly subjects had higher levels of antibody to phos- phocholine, but when isolated, this antibody did not opsonize any of the vaccine strains of pneumococci. These results suggest that alternative strategies are needed to maximize the protective effect of pneumococcal vaccine in the population at greatest risk. © 1986 Oxford University Press.