Field studies of grazing management have frequently concluded that the magnitude and direction of vegetation response is dependent on initial vegetation condition. On upland heath, this dependence reflects the importance of small-scale ecological processes (e.g. plant competition), and local neighbourhood effects (e.g. spatial distribution of plant species), in driving the vegetation dynamics. These small-scale effects, together with variation in grazing patterns, increase the difficulty of deriving general rules about the effect of grazing on vegetation change from field studies. However, we need to determine the impacts of such grazing-related vegetation change upon biodiversity, (e.g. birds). For many bird species it is impractical to use experimental approaches due to low breeding densities, and the influence of other site and management effects (e.g. predator control). To predict the effect of management changes on them requires an accurate assessment of the large-scale effects of grazing management on the ecological landscape using data from small-scale field studies. This paper sets out an approach that integrates field studies with theoretical models to investigate the large-scale effects of grazing management on plant and bird communities on upland heath.