There are scant data about normal reference values of blood glucose (BG) in children. This study was conducted to learn the BG profile of children and adolescents in Beijing area. The population for survey was selected as a stratified cluster sample from 8 urban and 10 rural areas in Beijing. Fasting capillary blood glucose (FCBG) was determined in 19,593 children and adolescents aged 6 to 18 years in 4 urban and 3 rural areas using haemosaccharometer model II [Roche Diagnostic, (Shanghai) Ltd]. There were 1 9112 (97.5%) individuals with complete records, the mean age was 12.1 +/- 3.3 years (ranged from 6 to 18.9 years); 9514 (49.8%) were boys, 9598 (50.2%) were girls, 9792 were (51.2%) from urban areas and 9320 (48.8%) from rural areas. The average level of FCBG in boys was higher than that in girls (4.7 +/- 0.5 vs. 4.5 +/- 0.5, u = 28.0, P < 0.01). Among urban children, the trend of variation of FCBG was similar between boys and girls, the levels of FCBG increased with age, the peak of FCBG was reached at 12-13 years in urban girls, and from the age of 15 years, the level of FCBG declined. In boys, the FCBG level increased slowly from 13 years of age, there was no significant variation until 17 years old, and declined at the age of 18. Among suburban children, the trend of variation of FCBG was similar between boys and girls, both of them had two peaks, from 6 to 11 years old, FCBG of both boys and girls increased with age, and both reached the first peak at the age of 11 years. While at 13 years of age, there was an obvious drop in FCBG level. From 14 years of age on, there was a rise of FCBG in both boys and girls, and the second peak of FCBG was reached at 15 and 16 years of age in girls and boys respectively. The FCBG level of urban children was higher than that of rural children (4.7 +/- 0.5 vs. 4.6 +/- 0.5, u = 13.8, P < 0.01). The level of FCBG in overweight and obese children was higher than that of normal children. More boys, more obese and more urban children had abnormal FCBG. The blood glucose level of children was associated with age, gender, obesity and district.