Neonatal hyperglycaemia is frequent and requires insulin therapy. To resolve the difficulties encountered by paediatricians in stabilising glycaemia, the preparation and administration of insulin aspart were assessed and optimised. After high-performance liquid chromatography (HPLC-UV) assessment of insulin aspart preparations made according to the old protocol, a new protocol was drawn up. Dosage reliability of solutions prepared by paediatric nurses was evaluated by HPLC-UV. This new protocol was also tested in a Y-infusion situation and the need to saturate infusion tubes assessed. Wide deviations in insulin aspart concentrations were found between theoretical concentrations and preparations made according to the old protocol. Glycated insulin aspart was found in the majority of these preparations. The new protocol significantly reduced the variability of data and relative deviations around the target value. It also eliminated the formation of glycated insulin even in the case of co-infusion of parenteral nutrition and confirmed the need to saturate infusion tubes. Conclusion: The revision of the insulin therapy protocol reduced the variability of insulin concentration in preparations and avoided the administration of glycated derivatives potentially toxic for neonates.