This paper evaluates the causal effects of working part-time and working full-time on body weight using data from the Health and Retirement Study. Additionally, we study these effects when we allow for heterogeneity across socio-economic status measured by income and education. Since retirement decision is a potential source of endogeneity, we follow an instrumental variable approach using retirement eligibility ages as instruments. We also investigate possible mechanisms which could explain the effects of working part-time and working full-time on body weight. To do this, we use data from the American Time Use Survey. We find that retirees have higher BMI than old workers. We also find that individuals with a higher socio-economic status usually have lower BMI. Furthermore, the allocation of time made by retirees and full-time workers on activities such as watching TV seems to favor sedentarism.