How soil microbial respiration responds to climate change can be constrained by historical climate. Understanding the duration of such legacy effects is key to determining how much they matter for projecting future ecosystem carbon cycling. Here, we tested whether extreme changes in rainfall could overcome constraints imposed by historical rainfall on how soil respiration responds to moisture. We predicted that larger shifts in rainfall regime would alter the magnitude or sensitivity (slope) of the respiration response to moisture compared to smaller changes in rainfall relative to historical conditions. Over 4.5 years, we imposed rain treatments ranging from extreme dry to extreme wet conditions that varied by ~400%, as well as ambient and historical mean rainfall controls. Rain treatments were applied to shortgrass or tallgrass vegetation that represented lower and higher biomass inputs, respectively, to test how shifts in soil resources might affect respiration moisture responses. We found high resistance to altered rainfall in the field, with persistent legacies indicated by no change in the respiration response to moisture among treatment and control rain treatments. The intrinsic respiration response to moisture under controlled laboratory conditions was also unaffected by field rain treatments. In response to field vegetation treatment, there was 10–30% more soil respiration in tallgrass compared to shortgrass that was paralleled by an increase in soil dissolved organic carbon, but no change in moisture sensitivity consistent with independent resource and climate effects. Soil bacteria and fungi were unchanged across all treatments and were largely generalists, suggesting high community as well as functional resistance to change. Climate legacies on soil microbial communities have the potential to modify our expectations for the rate of acclimation and adaptation to altered climate conditions.