This study aims to identify the association between air pollution and perinatal outcomes in the municipality of Santos, as well as to identify risk areas in the city for such an outcome from January 2012 to December 2015. The design is mixed, with parts of a retrospective cohort, as it works with secondary data, and ecological study, as it works with spatial data. Logistic regression analysis was performed with 10,319 live births. Live birth data were obtained through the Information System on Live Births (SINASC). Daily records of air pollutants (PM10, PM2.5, NO2 and O3), temperature and relative humidity were obtained through the São Paulo State Environmental Agency (CETESB). Spatial scan analysis was performed to identify areas of risk for prematurity and then, compared with census sectors according to the Paulista Index of Social Vulnerability (IPVS). Analysis were also performed to calculate the Moran Index to identify spatial correlation of cases between neighborhoods. The variables prenatal visits, mother's age and number of children born deceased showed a significant positive association with the outcome of LBW and prematurity. For Low birth weight: when pollutants were considered continuously, PM2.5 and PM10 and NO2 showed a protective effect in the third trimester, but with no association in the first and second, while O3 showed no association in any of the trimesters. For PM2.5, no significant association was found for the first and third trimester. Only the second quartile of the second trimester (OR = 1.41, 95% CI: 1.12 -1.76) showed a positive association with statistical significance. Only the last quartile of the third trimester (OR = 1.27 95% CI 1.23 -1.57) showed a positive association with significance. A significant association was found for the second and third quartiles in the second and third trimesters for PM10. For prematurity: no association was found between NO2 and PM2.5 and prematurity. O3 showed a significant association in the first trimester, in the last quartile (OR = 1.47 CI 95% 1.05; 2.07). PM10 showed a significant association in the first trimester, in the last quartile (OR = 1.28 CI 95% 1.00; 1.64) and in the second trimester, in the second quartile (OR = 1.37 CI 95% 1.07; 1, 77). Risk areas for prematurity and LBW had more census sectors with medium and high vulnerability. Areas with low risk for LBW and prematurity had census sectors with very low vulnerability. Results show evidence that maternal exposure to pollutants can contribute to prematurity and LBW, as well as suggest that having homes in a region of high vulnerability can also contribute to these outcomes.