on behalf of the IDEFICS consortium OBJECTIVES: The aim of this study is to present age-and sex-specific reference values of insulin, glucose, glycosylated haemoglobin (HbA1c) and the homeostasis model assessment to quantify insulin resistance (HOMA-IR) for pre-pubertal children. METHODS: The reference population consists of 7074 normal weight 3-to 10.9-year-old pre-pubertal children from eight European countries who participated in at least one wave of the IDEFICS ('identification and prevention of dietary-and lifestyle-induced health effects in children and infants') surveys (2007)(2008)(2009)(2010) and for whom standardised laboratory measurements were obtained. Percentile curves of insulin (measured by an electrochemiluminescence immunoassay), glucose, HbA1c and HOMA-IR were calculated as a function of age stratified by sex using the general additive model for location scale and shape (GAMLSS) method. RESULTS: Levels of insulin, fasting glucose and HOMA-IR continuously show an increasing trend with age, whereas HbA1c shows an upward trend only beyond the age of 8 years. Insulin and HOMA-IR values are higher in girls of all age groups, whereas glucose values are slightly higher in boys. Median serum levels of insulin range from 17.4 and 13.2 pmol l − 1 in 3-< 3.5-year-old girls and boys, respectively, to 53.5 and 43.0 pmol l − 1 in 10.5-< 11-year-old girls and boys. Median values of glucose are 4.3 and 4.5 mmol l − 1 in the youngest age group and 49.3 and 50.6 mmol l − 1 in the oldest girls and boys. For HOMA-IR, median values range from 0.5 and 0.4 in 3-< 3.5-year-old girls and boys to 1.7 and 1.4 in 10.5-< 11-year-old girls and boys, respectively. CONCLUSIONS: Our study provides the first standardised reference values for an international European children's population and provides the, up to now, largest data set of healthy pre-pubertal children to model reference percentiles for markers of insulin resistance. Our cohort shows higher values of Hb1Ac as compared with a single Swedish study while our percentiles for the other glucose metabolic markers are in good accordance with previous studies.