On average, specialized programs for pediatric overweight and obesity based on multicomponent behavioral interventions demonstrate efficacy. However, there is considerable heterogeneity in treatment response, particularly when considering attrition. Research on predictors of treatment participation and response suggests that many treatment failures are downstream consequences of broader social ecological factors. Advances in the availability of small-area spatial data on health outcomes has revealed large disparities in rates of pediatric overweight and obesity across neighborhood communities, suggesting a role for neighborhood‑level variables. A growing body of research has demonstrated that aspects of the built (physical human-made features) and social environment are associated with pediatric overweight and obesity and proximal lifestyle behaviors. There is also some research to suggest that these factors impact treatment participation and response, particularly among disadvantaged communities. However, few studies combine neighborhood-level predictors with clinical treatment outcomes. Moreover, most studies utilize additive regression methods that are not able to capture the complexity of neighborhood environments. Using Geographic Information Systems (GIS) the present study examined the home neighborhood environments of participants in a pediatric weight management program. Mixture modeling was used to characterize latent classes of neighborhood environments, in terms of patterns among built and social environment features, and to predict treatment participation and outcomes by class. Results revealed disparities in home neighborhood environments in terms of overall accessibility (of built environment features), relative accessibility, and social environment. In addition, these disparities were associated with early attrition and weight management outcomes in theoretically consistent ways.