A dynamic simulation model describing long-term forest development under different grazing pressures is presented as an illustration of system analysis in forest grazing research. In this paper, a brief description of the model is presented as are some simulation results. The FORest GRAzing model (FORGRA) simulates monthly changes in the characteristics of herbaceous understory and individual trees in plots, based on underlying physiological processes. Special attention is paid to the simulation of diet composition in relation to ungulate species, food quality and food availability, and accounting for seasonal variation in these factors. The model is parameterized for 12 tree and shrub species, six species of the ground flora and four ungulate species: roe deer (Capreolus capreolus), red deer (Cervus elaphus), Highland cattle (Bos taurus) and Konik ponies (Equus caballus). The results presented in this paper concern the development of a pine forest (Pinus sylvestris L.) in the Netherlands under various grazing pressures over a period of 100 years. They show that even low densities of ungulates can have significant impacts on the regeneration and thereby on forest development. It is argued that this approach is generally applicable in forest grazing research, providing a means for quantitative interpretation of the interaction between herbivory and forest development.