Background: The prevalence of obesity increased worldwide in children and adolescents from 1975 to 2016. The recent increase in childhood obesity has led to an interest in the question of which definitions should be used to distinguish the obese child. Body mass index (BMI) was recommended for use in children to assess body weight status. There are several international (World Health Organization, International Obesity Task Force, Center for Disease Control) and national BMI cut-offs references, and it is a major obstacle in studying global secular trends for younger age groups. Moreover, BMI does not distinguish between increased mass in the form of fat, lean tissue or bone, and hence can lead to significant misclassification. The ideal monitoring tool should directly assess adiposity. All available body composition methods in children are indirect. The gold standard for body composition is the four compartment (4-C) model. Although bioelectrical impedance analysis is most susceptible to imprecision when compared with the 4-C model it is the most logical bedside method to apply in children owing to its low cost, noninvasiveness, lack of radiation exposure and ease of use. Conclusions: BMI may produce a significant level of misclassification. Population-based cut-off values for body fat determined by body composition reference methods are the best criterion. Bioelectrical impedance is inexpensive, portable, simple and rapid to use. Further studies to elucidate the relationship among BMI, body fatness, fat distribution, and health risks in children should be followed.