Although multiple predator effects and trophic cascades have both been demonstrated in a wide variety of ecosystems, ecologists have yet to incorporate these studies into an experimental framework that also manipulates a common and likely important factor, spatial heterogeneity. We manipulated habitat complexity, the presence of two top predators (toadfish and blue crabs), and one intermediate predator (mud crabs) to determine whether habitat complexity influences the strength of multiple predator interactions across multiple trophic levels in experimental oyster reef communities. In the absence of toadfish, blue crabs caused significant mud crab mortality. Despite also directly consuming mud crabs, toadfish indirectly benefited this intermediate predator by decreasing blue crab consumption of mud crabs. Toadfish suppression of mud crab foraging activity, and thus decreased mud crab encounters with blue crabs, is likely responsible for this counterintuitive result. Contrary to previous investigations which suggest that more complex habitats reduce interference interactions among predators, reef complexity strengthened emergent multiple predator effects (MPEs) on mud crabs. The degree to which these MPEs cascaded down to benefit juvenile oysters (basal prey) depended on both habitat complexity and nonconsumptive effects derived from predator-predator interactions. Habitat complexity reduced the foraging efficiency of each crab species individually but released crab interference interactions when together, so that the two crabs collectively consumed more oysters on complex reefs. Regardless of reef complexity, toadfish consistently decreased consumption of oysters by both crab species individually and when together. Therefore, interactions between predator identity and habitat complexity structure trophic cascades on oyster reefs. Furthermore, these cascading effects of multiple predators were largely mediated by nonconsumptive effects in this system.