Climate modeling studies and observations do not fully agree on the implications of anthropogenic warming for evapotranspiration (ET), a major component of the water cycle and driver of irrigation water demand. Here we use California as a testbed to assess the ET impacts of changing atmospheric conditions induced by climate change on irrigated systems. Our analysis of irrigated agricultural and urban regions shows that warmer atmospheric temperatures have minimal implications for ET rates and irrigation water demands-about one percent change per degree Celsius warming (~1 %°C^-1). By explicitly modeling irrigation, we control for the confounding effect of climate-driven soil moisture changes and directly estimate water demand implications. Our attribution analysis of the drivers of ET response to global anthropogenic warming shows that as the atmospheric temperature and vapor pressure deficit depart from the ideal conditions for transpiration, regulation of stomata resistance by stressed vegetation almost completely offsets the expected increase in ET rates that would otherwise result from abiotic processes alone. We further show that anthropogenic warming of the atmosphere has minimal implications for mean relative humidity (<1.7%°C^-1) and the surface energy budget (<0.2%°C^-1), which are critical drivers of ET. This study corroborates the growing evidence that plant physiological changes moderate the degree to which changes in potential ET are realized as actual ET.