The selection of acellular vaccine antigens relies on current concepts of pertussis pathogenesis. Animal model data provide evidence that certain products of Bordetella pertussis, which include the putative adhesins filamentous hemagglutinin, pertactin and fimbriae, and pertussis toxin could serve as protective antigens and are available in sufficient quantities of purified material to be considered appropriate candidates for vaccine inclusion. In clinical studies vaccines containing three, four or five components were more effective at preventing pertussis than vaccines containing only inactivated pertussis toxin and filamentous hemagglutinin. These data suggest that pertactin may make a contribution to the protection elicited by an acellular product, but information does not allow evaluation of a possible incremental contribution from fimbriae. Serologic studies of patients in the clinical efficacy trials of the acellular pertussis vaccines did not yield a correlation between antibody levels and protection against pertussis, which suggests that relationships or mechanisms involved in the protective activities of these acellular vaccines are not yet understood. Therefore other mechanisms of immunity (i.e. cellular immunity) may be involved in vaccine-elicited immunity. Increasing understanding of the likely mechanisms of pertussis pathogenesis will provide insights into potential therapies for patients infected with B. pertussis. The mechanisms of vaccine-induced immunity remain elusive and determination of whether these products are working as initially predicted will require further study.