Type 1 diabetes (T1D) is an autoimmune disease whose incidence has been rising worldwide over the past three decades, suggesting that environmental factors may play a role in the onset of this condition, in addition to known genetic factors. Spatio-temporal variations in T1D incidence among children aged 6 months to 14 years during the 2010–2017 period in mainland France were described at the departmental level (n = 101, including the DROMs) and the cantonal level (n = 1,972, excluding the DROMs). Factors potentially correlated with spatial variations were investigated in a geographic ecological study at the canton level. An algorithm was used to identify new cases of T1D in the National Health Data System (SNDS) and to estimate the annual incidence over the period. Geographic indicators were constructed to characterize the cantons in terms of levels of chemical and physical pollutants, the presence of polluting activities, sociodemographic characteristics, or the incidence of certain risk factors associated with pregnancy. Associations between the cantonal incidence of T1D and cantonal characteristics were examined using Poisson regressions and hierarchical Bayesian models. A total of 16,683 new cases of T1D were detected from 2010 to 2017 (incidence rate = 17.2 per 100,000 children over the period, with a 4% annual increase). Spatial heterogeneity in T1D risk was identified, as well as a positive and statistically significant association between cantonal T1D incidence, levels of air pollutants (primarily NO₂), and other indicators of the canton’s level of social deprivation or urbanization. Cantonal levels of the other environmental pollutants studied were not associated with T1D incidence. This study provides a better understanding of the environmental characteristics associated with areas of elevated T1D incidence in children, including in particular characteristics linked to a high level of urbanization, and helps guide etiological research on the risk factors for this condition. It demonstrates the value of ecological approaches for integrating health and environmental data in the context of complex, multifactorial diseases, where risk factors (both environmental and sociodemographic) may be highly correlated with one another.