In many regions, forestry practices are shifting to partial harvesting approaches that seek to maintain species and structural diversity in managed forests. We monitored windthrow for 21 years following partial cutting treatments with 0%, 30%, and 60% removal in a large, replicated experiment located in mixed-species mature and old-growth forests of fire origin. There was no evidence that wind damage to merchantable trees (≥17.5 cm) varied among the three removal treatments. We found no evidence of a short-term spike in susceptibility to windthrow after partial cutting during the initial years following treatment. Over 21 years, a total basal area of 2.4 m2·ha–1 was damaged, which was 5.9% of the original standing basal area at the start of the experiment. We found clear differences in susceptibility to windthrow among the different tree species. The percentage of original standing trees that were windthrown varied from 0% to 23.7%. Eight of nine species had ≤10% damage over the monitoring period. Foresters should be aware of differences among tree species in risk of wind damage but should not use a general concern about susceptibility to windthrow as a reason to avoid partial cutting systems (that can achieve a diversity of management objectives) in structurally diverse, multispecies forests.