OBJECTIVES: Few studies have examined whether social characteristics of the residential environment are associated with blood pressure after controlling for individual sociodemographic characteristics. Even less is known about the processes by which these associations operate. Therefore, we examined whether distinct dimensions of the residential environment (socioeconomic position and urbanicity) were associated with systolic blood pressure. To better understand the processes involved in the associations between contextual factors and blood pressure, we assessed the extent to which these associations were mediated by body mass index and waist circumference. METHODS: We analysed data from the PRIME Study (7850, 50-60-year-old men surveyed in 1991-1993 in three French regions and recently geocoded on a local scale). We used multilevel regression models to estimate associations between contextual factors and blood pressure, and path analysis to investigate possible mediators of these associations. RESULTS: After adjustment for individual socioeconomic variables, systolic blood pressure increased independently with decreasing municipality population density and decreasing neighbourhood educational level. Path analysis indicated that approximately 37% of the association between neighbourhood education and blood pressure was statistically explained by the heavier weight and stronger central adiposity of people from deprived neighbourhoods. Approximately 19% of the association with population density was mediated by anthropometric factors. CONCLUSIONS: These data suggest that the neighbourhood environment may influence blood pressure; only part of the associations between contextual factors and blood pressure may operate through body weight and body shape modification.