Aim There is a high incidence of neonatal hypoglycaemia in neonates born to mothers with pre-existing diabetes. This often necessitates admission to the neonatal intensive care. Guidelines suggest maintaining intrapartum blood glucose levels (BGLs) of 4–7 mmol/l in women with diabetes to reduce the risk of neonatal hypoglycaemia. This study assessed whether intrapartum BGLs in women with pre-gestational Type 1 and 2 diabetes were predictive of neonatal hypoglycaemia. Methods A retrospective analysis of 261 births delivered at a tertiary hospital in Australia from 2009 to 2014. Results There were 122 cases of neonatal hypoglycaemia (glucose ≤ 2.6 mmol/l) in 261 births (47%). The mothers in the neonatal hypoglycaemia group spent less time with BGL in the range 4–7 mmol/l [55 ± 37% vs. 65 ± 35%, P = 0.02; odds ratio (OR) 0.992, P = 0.03] and more time with BGL in the 7–10 mmol/l range (31 ± 34% vs. 18 ± 27%, P = 0.003; OR 1.013, P = 0.003) compared with those without neonatal hypoglycaemia. Although statistically significant, receiver operating characteristic (ROC) curve analysis showed that time spent with maternal BGLs in the range 4–7 mmol/l [area under the curve (AUC) = 0.58] or 7–10 mmol (AUC = 0.60) was not strong enough to be a useful clinical predictor of neonatal hypoglycaemia. HbA1c in the second trimester of pregnancy (P = 0.02, OR 1.42) and percentage time spent in BGL range of 7–10 mmol/l (P = 0.001, OR 1.02) were both associated with a risk of neonatal hypoglycaemia in a logistic regression model. HbA1c in the third trimester (P = 0.07, OR 1.28) approached, but did not reach, significance. Conclusions These data support a BGL range of 4–7 mmol/l as an intrapartum target. Glycaemic control in the second trimester is associated with neonatal hypoglycaemia. Improvement in ante- and intrapartum glycaemic control may reduce neonatal hypoglycaemia in women with pre-existing diabetes.