The shelterwood system is considered appropriate to regenerate Pinus strobus. However, there is a need to quantify the amount of harvesting damage that can be expected relative to the amount of overstory removed during removal harvests and the amount of regeneration that exists prior to harvest. We thus evaluated regeneration response to harvest intensity, as expressed by the percent of basal area harvested. We tested whether regeneration quality and quantity post-harvest increase with decreasing harvest intensity and depend on their pre-harvest abundance. We also posited that soil disturbance increases with harvest intensity. We observed that declines in stocking and density were related to increasing harvest intensity, although most reductions were likely related to skid trails coverage. High pre-harvest densities helped offset some of the losses due to harvesting. A quarter of the regenerating trees alive post-harvest presented some form of damage; the occurrence of damages was not significantly influenced by the percentage of basal area harvested. To ensure regeneration objectives are met, losses during removal harvests should be accounted for; as harvesting intensity increases, losses also increase. Based on these results, minimizing skid trail coverage appears as the most effective way to reduce regeneration losses and damage during shelterwood harvests.