Objective To analyze the effects of previous history of maternal type 2 diabetes mellitus (T2D) in pregnant women and gestational diabetes mellitus (GDM) on the incidence of neonatal respiratory distress syndrome (NRDS), in order to provide evidence for preventing NRDS. Methods A retrospective cohort study was conducted among pregnant women hospitalized in Cangnan County People's Hospital from January 2015 to September 2020.Electronic medical records were retrospectively collected.Cox regression risk ratio of potential confounding factors was adjusted, then the relative risk of NRDS associated with maternal intrauterine diabetes in preterm infants was estimated. Results A total of 618 pairs of mothers and infants were included in this study, of whom 67 (10.8%) were exposed to T2D, 106 (17.2%) were exposed to GDM, and 80(12.9%) premature infants were diagnosed with NRDS.The prevalence of NRDS in infants exposed to T2D, GDM or no diabetes was 25.4% (17/67), 21.7% (23/106) and 9.0% (40/445), respectively.Risk analysis showed that newborns with gestational age of 32 to 34 weeks at delivery had a higher risk of NRDS than those at 35 to 37 weeks (HR=2.11, 95%CI: 1.21 - 3.96, P＜0.001).Compared with no diabetes during pregnancy, gestational diabetes with previous T2D (HR=1.96, 95%CI:1.28 - 3.42, P=0.006) and requiring antidiabetic drugs during pregnancy(HR=1.83, 95%CI: 1.22 - 3.10, P=0.017) were significantly associated with increased risk of NRDS.For gestational diabetes not requiring antidiabetic medication during pregnancy, there was no correlation between GDM and NRDS (P=0.375). Conclusions NRDS is predominantly attributed to exposure to the established T2D environment compared with non-diabetes during pregnancy.And the risk for NRDS is lower in children with maternal GDM requiring medication treatment, while the risk would not be significantly increased in children with maternal GDM not requiring medication.