Regression models to predict total tree height and maximum crown radius as a function of diameter at breast height were developed for Caribbean trees using data collected by the U.S. Forest Service in the Commonwealth of Puerto Rico and Territory of the U.S. Virgin Islands. Nonlinear models predicting height fit the data well with an overall pseudo-R² = 0.9220, as did models of height to diameter by Holdridge life zone and by the most frequently encountered species. The linear model predicting maximum crown radius for all trees combined fit the data poorly (R² = 0.3478). Crown model fits showed only moderate improvements when the data were modeled by species, crown class, and inventory measurement protocol, highlighting the variability of Caribbean forest tree crowns within and between species. Height models presented here will be useful for applications such as growth and yield simulation, forest health monitoring, and wildlife habitat modeling, but the crown radius prediction models only should be applied with an understanding of their limitations.