Herbivorous mammals can have profound effects on the sustainability and productivity of grasslands through their effects on nutrient cycling. Recent studies suggest that herbivores can accelerate the cycling of nitrogen by returning mineral nitrogen in urine and feces to the soil and favoring more nutrient-rich and decomposable plants. Alternatively, they can slow down nutrient cycling by modifying the species composition of plant communities, reducing litter quality, and slowing decomposition. However, an important question remains: what controls which of these effects herbivores will have? This LTER cross-site research project addresses six major hypotheses about what controls mammalian herbivore effects on nitrogen cycling in temperate grasslands. These hypotheses suggest that herbivore effects may vary in response to productivity, the prevalence of legumes in the plant community, and the susceptibility of soil to compaction. In addition, the density of different sized herbivores and their migratory versus seasonal use of grasslands may influence nitrogen cycling. Hypotheses will be tested by measuring herbivore impacts on nitrogen cycling within an existing cross-site study of the effects of different sized herbivores on grassland biodiversity and ecosystem function. These sites encompass a wide range of productivity, legume abundance, soil types, and patterns of seasonal use by herbivores. By comparing particular pairs of sites to test specific hypotheses and examining broad patterns in herbivore effects across all sites, this study will provide unique insights that are not possible with studies at a single site. The research will allow understanding of how different herbivore species affect the sustainable production of grasslands at sites with different productivity, plant species composition and soil types. Such information is critical to evaluating the role of mammalian herbivores, including domestic livestock, in the sustainable use and conservation of grasslands.