Large-scale manipulations of dead wood in mature Pinus taeda L. stands in the southeastern United States included a major one-time input of logs (fivefold increase in log volume) created by felling trees onsite, annual removals of all dead wood above P10 cm in diameter and P60 cm in length, and a reference in which no manipulations took place. We returned over a decade later to determine how these treatments affected tree growth using increment cores. There were no significant differences in tree density, basal area or tree diameters among treatments at the time of sampling. Although tree growth was consistently higher in the log-input plots and lower in the removal plots, this was true even during the 5 year period before the experiment began. When growth data from this initial period were included in the model as a covariate, no differences in post-treatment tree growth were detected. It is possible that treatment effects will become apparent after more time has passed, however. Published by Elsevier B.V.