Background/Question/Methods The relationship between biodiversity and ecosystem functioning (BEF) is central to ecology in theory and practice. Much of our current understanding of BEF and its underlying mechanisms is based on field manipulative experiments focusing on primary producers and carried out in artificial grasslands. This talk will introduce a large-scale, multi-trophic BEF experiment in the largest natural grassland in the world, the Inner Mongolia Grassland Removal Experiment (IMGRE). The overall goal of IMGRE is to test stoichiometric controls on the influence of plant functional diversity on ecosystem functioning across three trophic levels (plants, herbivores, and soil microbes), with an experimental design that compares and contrasts equal-disturbance and complete removal schemes. The project aims to synthesize core principles derived from recent advances in BEF and ecological stoichiometry (ES) research to clarify some of the ambiguities in functional diversity and niche complementarity. To achieve these goals, our research project is structured by four key hypotheses based on which several specific predictions are tested. Results/Conclusions In the past three years, we have measured ecosystem responses to the manipulations of plant functional diversity in terms of both the number of stoichiometrically-defined plant functional types (i.e., PFTs delineated by stoichiometric traits) and relative abundance. Ecosystem functioning variables include aboveground net primary production, C storage, N retention, N use efficiency, N mineralization, and other variables. While this talk focuses on the overall design and implementation of IMGRE, we will also discuss some of the challenges and lessons that we have learnt from such a large-scale BEF experiment, and present ongoing experimental manipulations with higher-trophic levels. In addition, several key findings so far from the project will be summarized to evaluate how well the overall research goal is being met. We expect the IMGRE project to contribute to the development of BEF theory by improving our understanding of stoichiometric mechanisms at multiple trophic levels. The research results will help improve the management practices of arid and semi-arid ecosystems, particularly the vast grasslands of the Eurasian Steppe Region.