The objective of the study was to evaluate the association of sociodemographic, biopsychosocial and lifestyle behaviors (diet, sedentarism and physical activity) with eating disorders (ED) in a national sample of Costa Rican children and adolescents. This was a cross-sectional and descriptive study. A total of 2667 students of elementary, middle and high school from sixty-four educative centers (40 schools and 24 high schools) of Costa Rica participated in the study. The predictive factors associated with ED are: body image dissatisfaction (β = 0.367; p < 0.001), lifetime addictive behaviors (β = 0.131; p < 0.001), to be a girl (β = 0.109; p < 0.001), intentionally hurt or harm itself (β = 0.074; p < 0.001), BMI (β = 0.053; p < 0.001), socioeconomic index (β = 0.052; p < 0.001) and sedentary behaviors (β = 0.051; p < 0.001). While the frequency of physical activity (β = −0.071; p < 0.001), the diet quality index-score (β = −0.068; p = 0.010) and the age (β = −0.064; p = 0.001) are not related with the development of ED. Psychosocial factors (body image dissatisfaction and intentionally self-injure) had the greatest influence (near 27%) of the global variance of the regression model that explained the association with ED. Programs should be promoted to monitor children and adolescents with excess weight, as well as, the coexistence of erroneous weight and body image perceptions in these population. This could protect against the development of ED in children and adolescents. Evidence-Based Medicine: Level V of evidence (descriptive study).