Background: HbA1c levels are infl uenced by the glycemic control of previous 2-3 months. Sometimes patients have surprisingly low HbA1c in spite of many correctly measured high blood glucose values, which is diffi cult to explain. As glucose sensors give an objective picture based on glucose readings several times per minute over 24 hours, we used the area under the curve (AUC) of such subcutaneous glucose profi les to evaluate their relationship with HbA1c. Methods: Thirty-two patients were randomized into two study arms, one open and the other blinded. Both arms had 8 pump users and 8 patients with multiple daily injections (MDI). After three months the two arms crossed over. Both study arms wore a continuous glucose monitoring system (CGMS) for 3 days every 2 weeks. HbA1c was determined before and after each 3-month study period. Results: There was no relationship between HbA1c and s.c. glucose AUC or between HbA1c and the number of peaks >15.0 mmol/L when all CGMS profi les during the 6 months were taken together. Children on MDI showed a positive relationship between HbA1c and AUC (P<0.01) as well as the number of peaks (P<0.01). Children with a negative relationship between HbA1c and AUC generally had fewer fluctuations in blood glucose values, whereas children with a positive relationship had wide fluctuations. Conclusions: Although there was no relationship between s.c. glucose AUC and HbA1c, the results indicate that wide blood glucose fluctuations may be related to high HbA1c values. Therefore, complications and therapeutic interventions should aim at reducing such fluctuations. © Springer 2008.