This study evaluates the effects of Mínimo Vital, a public policy implemented in Bogotá (Colombia) in 2012 that provided disadvantaged households with a free allowance of 6 cubic meters of water per month. In Colombia, urban housings are classified into socioeconomic strata on an ascending scale ranging from 1 to 6. Mínimo Vital was applied to strata 1 and 2 only, which allows one to provide quasi-experimental evidence on its effects. In particular, I combine a difference in differences identification strategy with entropy balancing, a method that allows one to estimate weights that perfectly balance observable characteristics across treated and control households. I estimate the causal effect of Mínimo Vital on the water bill, the probability of water reutilization, rainwater collection and the prevalence of fever and diarrhea in early childhood. As a result of the policy, I find an increase of USD$3.97, or 35%, in the water bill, an increase of 8.4 pp in the probability of water reutilization, and no effect on the probability of rainwater collection. Analysis by wealth index quartile suggests that the impacts of the policy are stronger for poorer households. Importantly, I also find that the prevalence of diarrhea among young children in the poorest households was significantly reduced by the policy.