ABSTRACT Obesity remains highly prevalent among adolescents and young adults in the United States and relatively few youth with overweight or obesity are able to lose weight and maintain the loss, thus making prevention of paramount importance. Although a range of genetic loci and behavioral and environmental factors predictive of weight change and obesity have been identified in prospective studies, their associations with weight gain and obesity have been modest in magnitude. Moreover, the numerous clinical trials aimed at prevention of obesity in children and adolescents have yielded modest effects. One possible reason for the lack of observing large effects in observational studies or clinical trials is that all types of overweight and obesity are often grouped together. One potential consequence of overlooking the heterogeneity in obesity is missing or underestimating risk factors for specific subtypes of obesity. An alternative strategy is to identify subtypes and examine subtype- specific associations. More careful phenotyping of obesity could potentially be used to advance the field of obesity epidemiology and treatment. Research is needed to identify the range of obesity phenotypes and whether there are differences in obesity subtypes by biological sex and race. We propose using appetitive behaviors, psychological, behavioral, anthropometric, physiological, genetic, and familial data, in a person-centered approach to empirically identify subtypes of obesity and to investigate whether the subtypes vary by biological sex or race. We will use data collected from the 2379 girls in National Growth and Health Study, 1421 children (48% with obesity and 17% with overweight) studied in the Section on Growth and Obesity at the Eunice Shriver Kennedy National Institute of Child Health and Human Development (NICHD), and the 16,850 youth in the Growing Up Today Study (14.9% with obesity and 29% with overweight). We will use latent class analysis to identify subgroups of obesity and then examine whether the empirically-derived phenotypes are predictive of weight change in adolescence and young adulthood, the development of hypertension before age 35, or have a differential pattern of association with the obesity loci. In addition, we will examine whether the phenotypes vary by biological sex or race and whether the associations of phenotypes with weight change in adolescence and young adulthood and the development of hypertension before age 35, vary by biological sex or race. The results of our proposed study could have a major impact on the field. The first step towards decreasing disparities by race and sex is to examine if there are differences in presentation and course of a disorder. Our study will provide that much needed information, which could result in tailored and more effective prevention and treatments.