AIM: The aim of this study was to identify the main contributors to cesarean section (CS) among women with and without diabetes during pregnancy using the Robson classification and to compare CS rates within Robson groups. METHODS: A population-based cohort study was conducted of all women who gave birth in New South Wales, Australia, between 2002 and 2012. Women with pregestational diabetes (types 1 and 2) and gestational diabetes mellitus (GDM) were grouped using the Robson classification. Adjusted odd ratios (AOR) and 95% confidence intervals (CI) were calculated using multivariable logistic regression. RESULTS: The total CS rate was 53.6% for women with pregestational diabetes, 36.8% for women with GDM and 28.5% for women without diabetes. Previous CS contributed the most to the total number of CS in all populations. For preterm birth, the contribution to the total was 20.5% for women with pregestational diabetes and 5.7% for women without diabetes. Compared to women without diabetes, for nulliparous with pregestational diabetes, the odds of CS was 1.4 (95% CI, 1.1-1.8) for spontaneous labor and 2.0 (95% CI, 1.7-2.3) for induction of labor. CONCLUSION: A history of CS was the main contributor to the total CS. Reducing primary CS is the first step to lowering the high rate of CS among women with diabetes. Nulliparous women were more likely to have CS if they had pregestational diabetes. This increase was also evident in all multiparous women giving birth. The high rate of preterm births and CS reflects the clinical issues for women with diabetes during pregnancy.