There is considerable disagreement about the impact of irrigation on heat stress, partly stemming from the choice of heat stress metric. Moreover, existing modeling studies, often used to examine impacts of irrigation, are done at scales that cannot appropriately resolve population centers, and thus potential for human impacts. For instance, most of us now live in urban areas, which locally modify heat stress, and thus may alleviate or aggravate irrigation-induced heat hazards. Here, we employ multi-year regional climate simulations at a convection-permitting scale to demonstrate that irrigation generally reduces summertime heat stress across more than 1600 urban clusters in North America. This holds true for most physiologically relevant heat stress metrics. The impact of irrigation varies by climate zone, with more notable irrigation signals seen for arid urban clusters in regions, which are near heavily irrigated fields. Using a component attribution framework, we show that the change in wet-bulb temperature, often used as a proxy for moist heat extremes in the geosciences, due to irrigation exhibits an opposite sign to the corresponding change in wet bulb globe temperature — a more complete metric for assessing both indoor and outdoor heat risk. Additionally, we explore the interplay between urbanization and irrigation effects on urban heat stress, highlighting the complexities in accurately interpreting their impacts. Of note, we use these results to discuss a spatiotemporal incoherence in how climate studies deal with hazard versus exposure and what that might mean for quantifying realistic heat-related risks for human populations. These findings also offer valuable insights into the role of local to regional land management practices on urban heat stress.