Forest disturbances, including whole-tree harvest, will increase with a growing human population and its rising affluence. Following harvest, forests become sources of C to the atmosphere, partly because wetter and warmer soils (relative to pre-harvest) increase soil CO2 efflux. This relationship between soil microclimate and CO2 suggests that climate changes predicted for the northeastern US may exacerbate post-harvest CO2 losses. We tested this hypothesis using a climate-manipulation experiment within a recently harvested northeastern US forest with warmed (H; +2.5 °C), wetted (W; +23% precipitation), warmed + wetted (H+W), and ambient (A) treatments. The cumulative soil CO2 effluxes from H and W were 35% (P = 0.01) and 22% (P = 0.07) greater than A. However, cumulative efflux in H+W was similar to A and W, and 24% lower than in H (P = 0.02). These findings suggest that with higher precipitation soil CO2 efflux attenuates rapidly to warming, perhaps due to changes in substrate availability or microbial communities. Microbial function measured as CO2 response to 15 C substrates in warmed soils was distinct from non-warmed soils (P < 0.001). Furthermore, wetting lowered catabolic evenness (P = 0.04) and fungi-to-bacteria ratios (P = 0.03) relative to non-wetted treatments. A reciprocal transplant incubation showed that H+W microorganisms had lower laboratory respiration on their home soils (i.e., home substrates) than on soils from other treatments (P < 0.01). We inferred that H+W microorganisms may use a constrained suite of C substrates that become depleted in their "home" soils, and that in some disturbed ecosystems, a precipitation-induced attenuation (or suppression) of soil CO2 efflux to warming may result from fine-tuned microbe-substrate linkages.