Spring snowmelt is the largest input to Arctic hydrological systems. The spring snow distribution is extremely spatially variable and difficult to quantify. This study used field measurements and models to characterize and quantify the spring snow distribution in the 52 km2 Niaqunguk River watershed near Iqaluit, Nunavut. Three models were assessed for their ability to replicate spatial patterns and estimate total watershed snow storage. Two semi-distributed terrain-based models were calibrated, and a fully distributed process model, SnowModel, was run. All 3 successfully replicated spatial patterns and provided reasonable quantitative estimates, except for SnowModel's poor performance in 2015. SnowModel is useful for studying mid-winter processes, but requires user technical capacity and high-quality meteorological observations lacking for much of the Arctic. By comparison, the semi-distributed models provide an accurate estimate without high technical or meteorological data demands, and provide a framework to guide stratified snow surveying.