Despite growing concern about the effect of declining environmental quality on human health, little effort has been made to quantify the effect of ecosystem degradation on the incidence and burden of infectious diseases. Using village-level administrative panel data and district-level satellite data on forest cover, I find that decline in forest cover from 2001-2008 in Indonesia can explain almost two million additional malaria infections. The evidence is consistent with an ecological response and the effect of deforestation on malaria cannot be explained by post-deforestation land use change, anti-malarial programs or migration. The effect is specific to malaria, with forest cover having no discernible effect on other diseases with disease ecologies different from that of malaria. Back of the envelope calculations suggest that the local health benefits of forest cover are greater than the global carbon benefits, underscoring a large, yet previously ignored and unquantified cost of forest loss with major implications for the design of payments for ecosystem services.