This dissertation investigates the causal link between early childhood health status and subsequent education outcomes. Its analytical approach focuses on the household as a decision making unit that invests in child health and education and takes into consideration those unobserved individual, household and community heterogeneities that might cause child health to be endogenous. The empirical analysis utilizes the fixed effects instrumental variables estimation method to obtain consistent parameter estimates of the effect of child health on education. The estimation overcomes the identification problem by using weather shock experienced during the birth year of a child to obtain an exogenous variation in child health. The estimates with household survey data from Vietnam reveal a significant effect of early childhood health status on schooling outcomes. The estimates suggest that considerable attenuation bias in the child health effect occurs when the endogeneity of child health is ignored. The estimates suggest that a nutrition intervention program will result in an empirically meaningful reduction in school entry delay and an improvement in schooling progress through grades.