It has been reported that the efficacy of pertussis vaccines appears to be lower when assessed in household contact studies than when assessed in cohort analyses of general populations. This observation may be attributable in part to methodological factors in home contact studies, in particular to the inclusion of retrospectively ascertained cases and of households in which the primary case constituted a vaccine failure. These biases are illustrated using data from a study of the efficacy of pertussis vaccines carried out by the Public Health Laboratory Service Epidemiological Research Laboratory PHLS/ERL in England during 1978-1980. This analysis reveals several factors which should be considered in the design, analysis and interpretation of household contact studies for evaluation of vaccine efficacy.