AbstractCan morphological plant functional traits predict demographic rates (e.g., growth) within plant communities as diverse as tropical forests? This is one of the most important next‐step questions in trait‐based ecology and particularly for global reforestation efforts. Due to the diversity of tropical tree species and their longevity, it is difficult to predict their performance prior to reforestation efforts. In this study, we investigate if simple leaf traits are predictors of the more complex ecological process of plant growth in regenerating selectively logged natural forest within the Wet Tropics (WTs) bioregion of Australia. This study used a rich historical data set to quantify tree growth within plots located at Danbulla National Park and State Forest on the Atherton Tableland. Leaf traits were collected from trees that have exhibited fast or slow growth over the last ~50 yr of measurement. Leaf traits were found to be poor predictors of tree growth for trees that have entered the canopy; however, for sub‐canopy trees, leaf traits had a stronger association with growth rates. Leaf phosphorus concentrations were the strongest predictor of Periodic Annual Increment (PAI) for trees growing within the sub‐canopy, with trees with higher leaf phosphorus levels showing a higher PAI. Sub‐canopy tree leaves also exhibited stronger trade‐offs between leaf traits and adhere to theoretical predictions more so than for canopy trees. We suggest that, in order for leaf traits to be more applicable to reforestation, size dependence of traits and growth relationships need to be more carefully considered, particularly when reforestation practitioners assign mean trait values to tropical tree species from multiple canopy strata.