Animal and human studies have shown an association between serum 25-hydroxyvitamin D (25(OH)D) level and insulin secretion and sensitivity. Accordingly, an association between 25(OH)D and glycated haemoglobin (HbA(1c)) is to be expected, and this was tested for in the present study. The Tromsø Study is a longitudinal population-based study initiated in 1974. In the sixth Tromsø Study conducted in 2007-2008, 12,984 subjects aged 30-87 years attended. After exclusion of current smokers and subjects with diabetes, the dataset consisted of 8643 subjects available for the present analyses. The correlation between serum 25(OH)D and HbA(1c) was -0.07 (p < 0.001). This association remained significant in a multiple linear regression model after adjustment for covariates gender, age, month of blood sampling, body mass index (BMI), physical activity score, serum triglycerides (TG), serum calcium and haemoglobin, and persisted across categories of gender, age, BMI and TG. The association appears to be most pronounced in the oldest, the obese and in those with the highest TG levels. Seasonal variation was found both for serum 25(OH)D and HbA(1c) with highest serum 25(OH)D levels and lowest HbA(1c) levels during summer months. In conclusion, there is a significant inverse association between serum 25(OH)D and HbA(1c) after adjustment for known confounders. The association is most pronounced in subjects with risk factors for glucose intolerance/type 2 diabetes. In a sub-analysis on subjects with diabetes the association between serum 25(OH)D and HbA(1c) appeared even stronger with a difference in HbA(1c) of 0.48 % between the highest and lowest serum 25(OH)D quartiles.