Scattered trees were uprooted, snapped, or bent during a 1983 windstorm in two northwestern Minnesota pine forests. I tested potential correlates (tree species, tree size, cause of damage) of damage type, and compared consequences of the uprooting, snapping, and bending of trees by assessing postdamage survival and by surveying formation and colonization of microsites (mounds, pits, stumps, and dead boles). Larger trees damaged directly by wind were usually snapped and killed in both study areas, regardless of tree species. Smaller trees damaged by falling neighbors were either (i) bent without sustaining mortality or forming microsites, a damage type prevalent among strong-wooded Acersaccharum and Ostryavirginiana in a Pinus–Acer study area, or (ii) uprooted, forming small mounds and pits, a damage type prevalent among weakwooded Abiesbalsamea and Picea spp. in a Pinus–Abies study area. Bent trees usually survived, at least for several years. However, uprooted and snapped trees were equally unlikely to survive their damage. Thus, uprooting and snapping were equally likely to result in dead bole formation. Windstorm-related microsites covered small proportions (6 and 18%) of the floor of the two forests, most as dead boles rather than as stumps, mounds, or pits. Microsites from uprooting, microsites from snapping, and background substrates did not differ in colonizing tree flora but did differ in density of colonists. The role of microsites varied with the autecology of tree species present. In the Pinus–Abies area, rotting wood of stumps and boles was the major establishment substrate for all regenerating tree species. In the Pinus–Acer area, Acersaccharum and A. rubrum were numerically dominant over other species on all substrates, with seedling densities highest away from microsites. Observed patterns of survival and tree regeneration indicate that uprooting and snapping of trees do not differ strongly in their consequences for either forest.