We compared, in 733 women with gestational diabetes mellitus treated with metformin and/or insulin, rates of neonatal hypoglycaemia in those who had received a dextrose/insulin infusion during labour and prior to delivery (n = 132) with those who did not (n = 601). Women who had infusions were more likely to have been treated with insulin (87.1% vs 70.4%, P < 0.01) and have higher mean capillary glucose values (measured four times daily) in the two weeks prior to delivery (P < 0.01). They had lower mean (SD) glucose values in the 12 h prior to delivery (5.1 (1.1) mmol/L vs 5.4 (0.9) mmol/L, P < 0.01). There was no difference between the groups in rates of neonatal hypoglycaemia (glucose <2.6 mmol/L on two or more occasions), 15.9% versus 17.8%, P = 0.78, or of severe neonatal hypoglycaemia (one or more glucose <1.6 mmol/L), 8.3% versus 5.2%, P = 0.15. In the absence of randomised data comparing use of infusions with no infusions, these data are reassuring for clinicians who do not routinely use infusions.