This paper formulates and estimates a structural intertemporal model of a woman’s household participating in the Supplemental Nutrition Assistance Program (SNAP) and her likelihood of being obese. We use an economic model of lifetime behavior in a finite life model to provide the structure of the econometric model, instrumental variable estimation is applied to control for endogeneity of SNAP participation decision, and individual fixed effects control for individual heterogeneity in panel data. Primary data are the panel, NLSY 79 with geocodes. We find that if a woman is in a SNAP household her BMI and probability of being obese are reduced by 15.7% and 56.3 percentage points, respectively. However, individual fixed effects account for much of the variation in her BMI and probability of being obese, suggesting early life attention to women’s weight is an important public policy issue.