To explore the sensitivity and specificity of waist to height index as indicator of overweight and obesity in pediatric age and to obtain cut-off points to simplify the diagnosis. Two thousand and three hundred and nineteen schoolchildren between 6 and 14 years were analyzed. Anthropometric measures were taken (height, weight, waist circumference and skinfolds thickness), and waist to height ratio (WHR), BMI and fat percentage were calculated. ANOVA test was used to evaluate the performance of anthropometric variables during the growth period. ROC curve analysis (Receiver Operating Characteristics) was applied using WHR as test variable and overweight and obesity status as criterion variable. Overweight and obesity were defined by BMI standards and adiposity references. The sensitivity and specificity values, areas under the curve (AUC), confidence intervals 95%, and cut-offs points were obtained. The statistical and graphical procedure was performed using SPSS(®) 18.0. WHR does not vary with age. AUC ranged from 0.786 to 0.953 indicating that the WHR has a high predictive power to identify the subjects previously classified as overweight or obese using both considered criteria. WHR proved to be an appropriate and effective predictor of overweight and obesity in children between 6 and 14 years. Cut-offs points of WHR that identify obesity are: 0.51 in males and 0.50 in girls. For the overweight, cut-offs range between 0.47 and 0.48 depending on sex and the criterion variable.