Abstract While one of the most striking features of our planet is its great variety of life, studies show that ongoing biodiversity loss could reduce the productivity of ecosystems by as much as 50%. However, evidence comes largely from experiments that have used highly simplified communities with on average seven species, all from a single trophic group. In contrast, natural communities have dozens, if not hundreds, of species spanning a variety of trophic levels. Would this additional complexity alter our conclusions about the functional consequences of diversity loss? This chapter reviews five hypotheses about how the fluxes of energy and matter through food-webs might depend on the diversity of species interacting within, as well as across trophic levels. After outlining the empirical support for or against each hypothesis, this chapter discusses several avenues of research that may prove useful as ecologists move towards a food web perspective on biodiversity and ecosystem functioning