The human-water feedbacks associated with agricultural water interventions may result in unintended and adverse externalities such as the phenomenon of supply-demand feedback, i.e. of increasing water demand in response to measures that enhance water supply. By integrating a spatially explicit hydrological model with a farmer’s behavioral module, simulated using Risks-Attitudes-Norms-Abilities-Self-regulation (RANAS) behavioral theory based on household surveys and rules derived from observed data, this paper emulates the phenomenon of supply-demand feedback. The Kamadhiya catchment in India is studied, where the decisions of 38,447 farmers in response to the intensive development of check dams (CDs) are simulated. The results show that the perceived increase in water supply from CDs has led nearby farmers to increase their cotton and wheat cultivation areas by 11.9% and 36.1%, respectively. This increase in demand has nullified the expected benefits from CD recharge with approximately 54.0% of the additional recharge being used for expanding irrigation for cotton and wheat, and lowering groundwater levels by 1.0 m. These findings underscore the need to critically understand human-water feedbacks that lead to unintended consequences and inform future investments in agricultural water interventions.