As the focus of forest management on many public lands shifts away from timber production and extraction to habitat, restoration, and diversity-related objectives, it is important to understand the long-term effects that previous management activities have on structure and composition to better inform current management decisions. In this paper, we analyzed 40years of growth data to quantify (1) the long-term response of yellow-poplar to thinning across an age and site quality gradient, and (2) the longevity of any growth response yellow-poplar may have to thinning throughout the southern Appalachian Mountains. Between 1960 and 1963, 134–0.1ha plots were established across an age and site quality gradient in yellow-poplar (Liriodendron tulipifera L.) stands throughout the southern Appalachian Mountains. All plots were thinned from below, with post-thinning relative density categorized into three classes: low (relative density <0.25), moderate (relative density ⩾0.25 but <0.35), and high (relative density ⩾0.35 but <0.60). Using plot-level annual basal area increment (BAI; cm2 yr−1) chronologies reconstructed from tree cores, average annual BAI was calculated for 10years prior to thinning (BAIpre) and each following 10year period thereafter (BAIpost).Site index and age at the time of thinning had a positive effect on BAIpost. During the first 10year period following thinning, annual BAI (at a site index=32.3m and age=43) averaged (SE) 33.7 (1.6), 26.3 (1.3), and 21.6 (1.1) cm2 yr−1 in the low, moderate, and high density classes, respectively. Significant differences between low and moderate and low and high density classes remained throughout the duration of the study. During the 10years post-thinning the ratio of BAIpost to BAIpre (RBAI) was >1.0 in 92%, 86%, and 57% of plots in the low, moderate, and high density classes, respectively indicating an overall increase in growth relative to pre-thinning growth rates. By the fourth decade post-thinning the percentage of stands containing trees that possessed RBAI values >1.0 had fallen; however trees in 71% of the plots in the low density class continued to experience growth rates greater than those prior to thinning. We conclude the increase in growth is short-lived when density is reduced to moderate and high levels whereas the response of trees to more intense thinnings is long-lasting.