Dietary and physical activity behaviors formed early in life can increase risk for childhood obesity and have continued negative consequences for lifelong health. Previous research has highlighted the importance of both genetic and environmental (e.g., cultural environment or parental lifestyle) contributions to obesity risk, although these studies typically involve genetically-related individuals residing in the same household, where genetic similarity and rearing environment are inextricably linked. Here we utilize a sibling-adoption design to independently estimate genetic and environmental contributions to obesity risk in childhood and describe how these influences might vary as children age. As part of a prospective adoption study, the current investigation used data from biological siblings reared either apart or together, and nonbiological siblings reared together to estimate the contributions of genetics and environment to body mass indices (BMI) in a large cohort of children (N = 711). We used a variance partitioning model to allocate variation in BMI to that which is due to shared genetics, common environment, or unique environment in this cohort during middle childhood and adolescence. We found 63% of the total variance in BMI could be attributed to heritable factors in middle childhood sibling pairs (age 5–11.99; 95% CI [0.41,0.85]). Additionally, we observed that common environment explained 31% of variation in BMI in this group (95% CI [0.11,0.5]), with unique environment and error explaining the remaining variance. We failed to detect an influence of genetics or common environment in older sibling pairs (12–18) or pairs spanning childhood and adolescence (large sibling age difference), but home type (adoptive versus birth) was an important predictor of BMI in adolescence. The presence of strong common environment effects during childhood suggests that early interventions at the family level in middle childhood could be effective in mitigating obesity risk in later childhood and adolescence.