Few other studies have investigated the association between deaths in the perinatal period and air pollution (Hajat et al. 2007; Lin et al. 2004; Nishioka et al. 2000; Pereira et al. 1998). Most studies have used a time-series approach and therefore examined the effect of air pollution on day(s) before the event (temporal scale). In addition, they relied on air quality monitoring sites for assessment of exposure to urban air pollution. In general, they all found statistically significant associations between daily counts of deaths (late fetal, neonatal, infant) and the average concentration of pollutants in the atmosphere. Pereira et al. (1998), in a study conducted in São Paulo, also showed that carboxyhemoglobin levels of blood collected from the umbilical cord of nonsmoking mothers were correlated with environmental carbon monoxide. These studies, although not directly comparable with the present study, which we based on spatial comparisons, suggest that fetuses and newborns may suffer from the consequences of a contaminated environment.