Abstract The spatial heterogeneity of mountain snow cover and ablation is important in controlling patterns of snow cover depletion (SCD), meltwater production, and runoff, yet is not well-represented in most large-scale hydrological models and land surface schemes. Analyses were conducted in this study to examine the influence of various representations of snow cover and melt energy heterogeneity on both simulated SCD and stream discharge from a small alpine basin in the Canadian Rocky Mountains. Simulations were performed using the Cold Regions Hydrological Model (CRHM), where point-scale snowmelt computations were made using a snowpack energy balance formulation and applied to spatial frequency distributions of snow water equivalent (SWE) on individual slope-, aspect-, and landcover-based hydrological response units (HRUs) in the basin. Hydrological routines were added to represent the vertical and lateral transfers of water through the basin and channel system. From previous studies it is understood that the heterogeneity of late winter SWE is a primary control on patterns of SCD. The analyses here showed that spatial variation in applied melt energy, mainly due to differences in net radiation, has an important influence on SCD at multiple scales and basin discharge, and cannot be neglected without serious error in the prediction of these variables. A single basin SWE distribution using the basin-wide mean SWE ( SWE ‾ ) and coefficient of variation (CV; standard deviation/mean) was found to represent the fine-scale spatial heterogeneity of SWE sufficiently well. Simulations that accounted for differences in ( SWE ‾ ) among HRUs but neglected the sub-HRU heterogeneity of SWE were found to yield similar discharge results as simulations that included this heterogeneity, while SCD was poorly represented, even at the basin level. Finally, applying point-scale snowmelt computations based on a single SWE depth for each HRU (thereby neglecting spatial differences in internal snowpack energetics over the distributions) was found to yield similar SCD and discharge results as simulations that resolved internal energy differences. Spatial/internal snowpack melt energy effects are more pronounced at times earlier in spring before the main period of snowmelt and SCD, as shown in previously published work. The paper discusses the importance of these findings as they apply to the warranted complexity of snowmelt process simulation in cold mountain environments, and shows how the end-of-winter SWE distribution represents an effective means of resolving snow cover heterogeneity at multiple scales for modelling, even in steep and complex terrain.