Rain-on-snow (ROS) events are becoming increasingly frequent in boreal regions, where they can trigger flooding and strongly influence seasonal snowpack evolution. The amount of energy available for melt is a key factor governing the hydrological impact of these events. However, the influence of forest canopy structure on individual energy fluxes, their relative contributions, and their seasonal variability remains poorly documented. In this study, we combined direct energy-flux measurements with estimation methods that account for canopy structure to analyze the snowpack energy balance during ROS events at three sites in the Montmorency Forest, Québec (47° N, 71° W): a clear-cut, a small forest gap, and a dense forest canopy. Over two winters (2023–24 and 2024–25), 20 ROS events were monitored. At the clear-cut site, turbulent heat fluxes dominated, contributing roughly 50% of the snowpack energy inputs, while net radiation accounted for about 30%. The relative contributions of the different fluxes remained generally stable throughout winter, except for enhanced ground heat flux during early-season events and increased shortwave radiation contributions toward late winter. At the gap and sub-canopy sites, reduced wind strongly limited turbulent exchange, and net radiation became the dominant component, contributing about 50% of the total energy input. The reduction in turbulent fluxes resulted in 55–60% lower energy inputs relative to the clear-cut site. Validation of energy-balance-derived snow mass changes against observations revealed a slight underestimation at the gap and sub-canopy sites, reflecting the limited applicability of conventional turbulent-exchange parameterizations under low-wind forest conditions.