Energy balance, net radiation and air temperature based models of daily snowmelt in a clearcut and below a forest were calibrated during April to June 1997. They were evaluated by comparison with snowmelt calculated using changes in snow depth from a snow depth sensor and measured snow density, and outflow from a snowmelt lysimeter (clearcut only) for April to June for 1998, 1999, and 2000. The sites are at 1630 m on the interior plateau, in southern British Columbia, Canada. In the clearcut, there was good agreement between all the modelled values and the measured data. The full energy balance calculation improved the ability to predict melt over only net radiation under warm and windy conditions where melt energy was doubled by the sensible and latent terms. The air temperature model worked remarkably well once the pack was ripe. Thus, this model should only be applied when the snow temperature is close to 0C. Good agreement was also obtained between calculated and simulated melt for the forest though the models tended to underestimate total melt.