We developed and evaluated an individual-tree height growth model for Douglas-fir [Pseudotsuga menziesii (Mirbel) Franco] in the Inland Northwest United States. The model predicts growth for all tree sizes continuously, rather than requiring a transition between independent models for juvenile and mature growth phases. The model predicts the effects of overstory and understory vegetative competition on height growth. Our model requires attained height rather than tree age as a predictor variable, thereby avoiding the problems of site index. Site effects are introduced as a function of ecological habitat type, elevation, aspect, and slope. We used six data sets totaling 3,785 trees in 314 plots. The structure of the data and the model indicated the need for a mixed-effects, nonlinear modeling approach using maximum likelihood in a linear differential equation with a power transformation. Behavior of the model was analyzed using a state-space approach. Our results show that both overstory and understory density affect height growth, allowing a manager to make informed decisions about vegetation control.