OBJECTIVE: infants of women with diabetes in pregnancy are at increased risk of hypoglycaemia. If the infant's blood glucose is low and the mother is unable to breast feed/provide sufficient expressed breast milk, infants are often given formula. Some hospitals encourage women with diabetes to express breast milk before birth. However, there is limited evidence for this practice, including its impact on labour and birth, e.g. causing premature birth may be a concern. A pilot study was undertaken to establish the feasibility of conducting an adequately powered randomised controlled trial to evaluate this practice. DESIGN: consecutive eligible women with pre-existing or gestational diabetes (requiring insulin), planning to breast feed and attending the study hospital were offered participation. INCLUSION CRITERIA: 34-36 weeks of gestation; singleton pregnancy; cephalic presentation; and able to speak, read and write in English. EXCLUSION CRITERIA: history of spontaneous preterm birth, antepartum haemorrhage, placenta praevia and suspected fetal compromise. Women were encouraged to express colostrum twice a day from 36 weeks of gestation, and advised how to store the colostrum, which was frozen for their infant's use after birth. They were asked to keep a diary documenting their expressing. DATA: demographic questionnaire, telephone interview at six and 12 weeks postpartum and medical record data. SETTING: a public, tertiary, women's hospital in Melbourne, Australia. PARTICIPANTS: 43 women with diabetes in pregnancy (requiring insulin). FINDINGS: cardiotocographs were undertaken after the first expressing episode and none of the infants showed any sign of fetal compromise. Forty per cent of infants received formula in the 24 hours postpartum. The proportion of infants receiving any breast milk at six weeks was 90%, and this decreased to 75% at 12 weeks. No women showed evidence of hypoglycaemia post expressing. The intervention was positively received by most women; 95% said that they would express antenatally again if the practice proved to be beneficial. The amount of colostrum varied according to the number of expressions, the length of time in the study and the time spent expressing, with a median of 14 days expressing and 39.6 ml of colostrum obtained. KEY CONCLUSIONS: the small number of women in this pilot was not an adequate number to examine safety or efficacy, but this study does provide evidence that it would be feasible and desirable to conduct a randomised controlled trial of antenatal milk expressing for women with diabetes requiring insulin in pregnancy. IMPLICATIONS FOR PRACTICE: it is important that this widespread practice undergoes rigorous evaluation to assess both efficacy and safety. Until such evidence is available, the authors suggest that the routine encouragement of antenatal milk expressing in women with diabetes in pregnancy should cease.