Objective: To determine whether the association between family history of diabetes (FHD) and impaired fasting glucose (IFG) is independent of body mass index (BMI) in children and adolescents. Methods: In all 443 (11.9%) children and adolescents with FHD, and 3280 (88.1%) without FHD were enrolled in a population-based cross-sectional study. Eligible subjects to participate were apparently healthy children and adolescents aged 7-15 yr from Middle and Northern Mexico. Obesity was defined by age- and gender-specific BMI >or= 95th percentile. FHD was defined as positive if at least one first degree relative had diabetes. The IFG was defined by fasting plasma glucose >or= 100 mg/dL and < 126 mg/dL. Results: IFG was identified in 390 (88.0%) and 62 (1.9%) children and adolescents with and without FHD, respectively. In the group with positive FHD, IFG was diagnosed in 146 (37.4%), 79 (20.2%), and 165 (42.3%) children and adolescents who were obese, overweight, and normal-weight, respectively. On the other hand, in the group without FHD, IFG was identified in 21 (33.9%), 14 (22.6%), and 27 (43.5%) children and adolescents who were obese, overweight, and normal-weight, respectively. In the overall population, the age-, sex-, and BMI-adjusted logistic regression analysis showed a strong and independent association between FHD and IFG [odds ratio (OR) -11.7; 95% CI 9.5-21.2]. This association remained strong for girls and boys in a subsequent analysis stratified for BMI category. Conclusions: The presence of FHD in a first degree relative is associated with IFG, even in the absence of obesity.