Modeling forest trees’ mensurational relationships has many applications: from understanding physiological functional relationships (e.g. the allometric relationship of sapwood to leaf area ratio) to predicting one tree part by measuring another (e.g. above-ground tree biomass from the tree stem diameter). Simple models that relate easy to measure tree characteristics, such as diameter at breast height (DBH), to other more difficult to observe dimensions such as crown volume, are useful both in and of themselves and also as components of larger forest growth and yield and forest ecosystem models. Mensurational models of wide applicability include tree height-DBH (growth and yield models, carbon accounting, forest dynamics), bark thickness-DBH (fuels and fire models, carbon accounting, post-fire mortality), and crown dimensions-DBH (physiological models, growth and yield models, carbon accounting models). Here we develop models for plantation-grown loblolly pine (Pinus taeda L.) trees in Uruguay. We model total tree height-DBH, crown diameter-DBH, crown volume-DBH, and bark thickness stem profile given relative diameter and relative height. We do so using linear and nonlinear regression models as well as semi-parametric modeling alternatives. The data comprise 198 trees grown in even-aged plantations and spanning an age range of 8–22years. The predictive ability of the resulting models was evaluated by estimating the root mean square prediction error by predictor variable classes with the 0.063+ bootstrap method. All models presented acceptable prediction ability except the one describing the relationship of crown volume to DBH. The semi-parametric model describing the bark thickness profile along the tree stem fit the data better and had similar prediction ability compared with the parametric model. The semi-parametric modeling approach was a good alternative to describe this allometry.