Hypoglycaemia is a major cause of neonatal morbidity and may induce long-term developmental sequelae. Clinical signs of hypoglycaemia in neonatal infants are unspecific or even absent, and therefore, precise and accurate methods for the assessment of glycaemia are needed. Glycaemia measurement in newborns has some particularities like a very low limit of normal glucose concentration compared to adults and a large range of normal haematocrit values. Many bedside point-of-care testing (POCT) systems are available, but literature about their accuracy in newborn infants is scarce and not very convincing. In this retrospective study, we identified over a 1-year study period 1,324 paired glycaemia results, one obtained at bedside with one of three different POCT systems (Elite? XL, Ascensia? Contour? and ABL 735) and the other in the central laboratory of the hospital with the hexokinase reference method. All three POCT systems tended to overestimate glycaemia values, and none of them fulfilled the ISO 15197 accuracy criteria. The Elite XL appeared to be more appropriate than Contour to detect hypoglycaemia, however with a low specificity. Contour additionally showed an important inaccuracy with increasing haematocrit. The bench analyzer ABL 735 was the most accurate of the three tested POCT systems. Both of the tested handheld glucometers have important drawbacks in their use as screening tools for hypoglycaemia in newborn infants. ABL 735 could be a valuable alternative, but the blood volume needed is more than 15 times higher than for handheld glucometers. Before daily use in the newborn population, careful clinical evaluation of each new POCT system for glucose measurement is of utmost importance.