Noy-Meir's simple but insightful model of grazing-system dynamics was used to draw broader inferences from empirical data generated by a 17-year field trial with beef cattle grazing a Mediterranean grassland in northern Israel. After calibration of its parameters against the field results, the model predictions were tested against an independent set of data obtained from the study site; they were within acceptable deviations from the inherently noisy field data. The calibrated model was used to analyse the effects of changes to two key grazing-management factors – stocking density and early-season grazing deferment – on biomass dynamics and forage consumption. The simulated results were used to calculate forage deficits and supplementary feed requirements for optimum herd performance during the growth (‘green’) season and throughout the year. The results revealed a critical stocking density of 0.7 Animal Units (AU)/ha, above which early-season deferment reduced the amount of supplementary feed required to maintain the optimum production of the herd. Optimum stocking is higher when the grassland is used mainly in the highly nutritious green season. Responses of the strongly seasonal Mediterranean grassland to the interaction between stocking density and early-season grazing deferment were expressed by a calibrated model, in terms that determine the efficiency of forage supplementation of the herd during the green season and throughout the year.