Two machine-learning techniques, gradient boosting (GB) and random forests (RF), were used to predict stand mean height (HT), trees per hectare (Tree ha-1) and basal area per hectare (BA ha-1) based on datasets collected from extensively- and intensively-managed loblolly pine plantations in the West Gulf Coastal Plain region. Models were evaluated using coefficient of determination (R2), bias and root mean squared error (RMSE) by applying models to independent dataset and then compared to the model (Coble et al. 2017) currently being used in the region. For extensively-managed plantations, the GB models had less bias, larger R2 and smaller RMSE than RF and HT model was the best, followed by those of Tree ha-1 and BA ha-1. Even for BA ha-1, the GB model had R2 over 0.83. GB and RF models outperformed the Coble et al. (2017); differences were notable for HT and Tree ha-1, but significant for BA ha-1. For intensively-managed plantations, GB and RF were similarly great in predicting HT and Tree ha-1, but GB outperformed RF in predicting BA ha-1. We recommend the use of GB models to predict quantitative information required for managing loblolly pine plantations in the region.