Three prediction methods were tested against observed melt for their ability to predict snowmelt in a variety of subarctic environments. Four melt environments were defined according to vegetation cover. They are: Closed Lichen Woodland, Open Lichen Woodland, Regenerating Burn and Burn. The three methods tested are the physical energy balance method (with some data approximations), the U.S. Army Corps of Engineers method (empirical energy balance) and the temperature index method (using mean daily air temperature). It was found that the physical energy balance is most applicable in the Burn and Regenerating Burn while the U.S. Army Corps of Engineers and temperature index methods provide adequate prediction in all four melt environments.