The authors' objective was to determine whether residential proximity to an industrial park (IP) is associated with increased perinatal mortality (PM). This semiecological study included 63,850 delivered births with 840 cases of PM (1995-2000). The authors categorized the study populations by ethnicity (ie, Bedouin and Jewish) and type of locality. Residential distance from the IP served as a surrogate indicator of exposure. Among Bedouin newborns, proximity to the IP was associated with increased PM rates (relative risk = 1.45; 95% confidence interval = 1.22-1.72). The excess in PM was not related to maternal or newborn physical characteristics that the authors observed. The risk of PM and its components in the Jewish localities was not associated with IP proximity. The association between residential proximity to the IP and excess in PM among only Bedouin newborns may be related to vulnerability caused by the nomadic nature of the society.