Data from nearly 11000 deliveries that occurred between 2005 and 2009 at the Besançon or Dijon (France) were collected during the PreCEE program, including medical, obstetrical, socio-economic and environmental conditions (noise, air pollution, proximity to green spaces). Several approaches were conducted to investigate the relationships between the occurrence of adverse pregnancy outcomes (APO) and the urban environment of two middle-sized cities, and to estimate the potential modulation of vulnerability on the impact of noise or air pollution during pregnancy. The mean NO₂ level ranged between 24 and 25 μg/m³ and the mean L_night ranged between 51 and 52 dB(A). The correlation between noise and NO₂ indices ranged from 0.41 to 0.59. The accumulation of 15 vulnerability markers regrouped in six dimensions (maternal age, smoking, body mass index, socio-economic, medico-obstetrical and environmental vulnerabilities) was analysed in comparison with four APO. Among the 3686 women included in this approach, 21% were not exposed to any marker and 19% accumulated three or more markers among the six dimensions. The risk of APO increased significantly with the cumulative number of vulnerabilities. Secondly, two case-control studies were conducted on preterm delivery on single pregnancies. First, only pregnancies without any associated comorbidities were considered (N=1511). No significant differences in pollutant exposure levels were found between cases and controls in this study. The second study was focused on pregnancies with associated comorbidities (N=2503). The OR associated with NO₂>40μg.m^-3 was 0.83 (95%CI: 0.61-1.13) and the OR associated with L_night&_geq;55dB(A) was 1.23 (95%CI: 0.97-1.56). A strong and monotonic association between the cumulative number of comorbidities during pregnancy and preterm delivery was also identified. Low levels of air pollution do not seem to affect preterm delivery. However, moderate noise exposure could potentially affect pregnancy when associated with vulnerability factors.