Purpose Recent studies suggest that body mass index is not a reliable enough measurement for body composition in individuals, particularly in older and younger people. However, most research on body image has used the body mass index (BMI) as a physiological predictor of body satisfaction, particularly in children. The aim of this study was to investigate whether body composition is a better predictor of body size dissatisfaction in children than BMI. Methods This is a cross-sectional study. Healthy children and adolescents aged 5-19 years, sex male and female, were recruited using a convenience sample in Brazil. BMI was measured according to the international standardization method and body composition was measured by bioelectrical impedance analysis (BIA) with a portable device model (BIA InBodyS10 multi-frequency, USA). Body size dissatisfaction was assessed using the Kakeshita's Figure Rating Scale for Brazilian Children. Data were analyzed with logistic regression analysis. Results A total of 547 participants were evaluated, including 54% females and 67% Caucasian, with a mean age of 11.4 ± 3.8 years. The mean BMI was 20.5 ± 4.6 kg/m 2 , and the mean percentages of fat and lean mass were 23.01 ± 10.59% and 72.84 ± 10.03%, respectively. In the multivariable model, only body composition was significantly associated with body size dissatisfaction (odds ratio: 1.849 (1.085-3.149, p = 0.024) and 1.828 (1.043-3.202, p = 0.035), respectively). Conclusions Body composition measures can better predict body size dissatisfaction in children and adolescents than BMI. This result may be relevant for the design of future studies on physiological indicators and body satisfaction. Level of evidence Level V, cross-sectional study