The marginal propensity to spend on food out of food stamps (MPSFS) is a key indicator of the fungibility of food stamp income. This fungibility has important implications for food stamp policy. Studies estimating the MPSFS use very different methods and come to very different conclusions regarding how households use food stamps. We use meta-analytic methods to investigate sources of heterogeneity in the previous MPSFS estimates. Our analysis describes the previous literature, and highlights the impact that study design characteristics have on the size of the MPSFS and therefore conclusions regarding household use of food stamp income.