In semi-arid mountainous regions across the western United States, the distribution of upland aspen (Populus tremuloides) is often related to heterogeneous soil moisture subsidies resulting from redistributed snow. As temperatures increase, interactions between decreasing snowpack and future trends in the net primary productivity (NPP) of aspen forests remain uncertain. This study characterizes the importance of heterogeneously distributed snow water to aspen communities in the Reynolds Creek Critical Zone Observatory located in southwestern Idaho, USA. Net primary productivity of three aspen stands was simulated at sites spanning elevational and precipitation gradients using the biogeochemical process model Biome- BGC and precipitation data adjusted to account for drifting snow. Compared to a spatially homogeneous precipitation distribution, Biome- BGC simulations accounting for redistributed precipitation were in better agreement with previous simulations of snow accumulation and soil moisture field measurements. During drought years, simulations below the largest drifts that included wind-redistributed snow resulted in NPP values nearly 77% higher than simulations assuming uniform precipitation. However, during wet years (and at sites with higher total precipitation), increased effective precipitation resulting from drifting snow did not have a significant role in aspen productivity. In these cases, soil moisture was found to be non-limiting even in the absence of redistributed snow. Increased water availability from snow drifts often exceeded the storage capacity of the soil and contributed little to plant available water used later in the growing season.