This work attempts to disentangle the relationship between credit, food security, and dietary diversity in the context of Bangladesh through descriptive and econometric analyses of the Household Income and Expenditure Survey of 2010 as well as a supplementary primary survey of 1200 households. To adequately address potential selection bias, we apply a variant of propensity score matching as well as an instrumental variable technique based on the distance from the nearest financial institution to account for endogeneity in our estimates. Our analysis reveals that access to credit tends to improve food security and allows households to achieve greater dietary diversity. In particular, food security is proxied by calorie consumption, and households with credit access tend to have greater calorie consumption per capita. Dietary diversity is measured through a number of dietary diversity scores, such as the food consumption score and the household dietary diversity score, and households with access to credit score higher than those without according to such measures. The results have been found to be robust following correction for endogeneity issues, and the paper therefore provides empirical evidence in favor of policies supporting accessible credit for poor households in Bangladesh.