Top-down control can be an important determinant of ecosystem structure and function, but in oceanic ecosystems, where cascading effects of predator depletions, recoveries, and invasions could be significant, such effects had rarely been demonstrated until recently. 2. Here we synthesize the evidence for oceanic top-down control that has emerged over the last decade, focusing on large, high trophic-level predators inhabiting continental shelves, seas, and the open ocean. 3. In these ecosystems, where controlled manipulations are largely infeasible, 'pseudo-experimental' analyses of predator-prey interactions that treat independent predator populations as 'replicates', and temporal or spatial contrasts in predator populations and climate as 'treatments', are increasingly employed to help disentangle predator effects from environmental variation and noise. 4. Substantial reductions in marine mammals, sharks, and piscivorous fishes have led to mesopredator and invertebrate predator increases. Conversely, abundant oceanic predators have suppressed prey abundances. Predation has also inhibited recovery of depleted species, sometimes through predator-prey role reversals. Trophic cascades have been initiated by oceanic predators linking to neritic food webs, but seem inconsistent in the pelagic realm with effects often attenuating at plankton. 5. Top-down control is not uniformly strong in the ocean, and appears contingent on the intensity and nature of perturbations to predator abundances. Predator diversity may dampen cascading effects except where nonselective fisheries deplete entire predator functional groups. In other cases, simultaneous exploitation of predator and prey can inhibit prey responses. Explicit consideration of anthropogenic modifications to oceanic foodwebs should help inform predictions about trophic control. 6. Synthesis and applications. Oceanic top-down control can have important socio-economic, conservation, and management implications as mesopredators and invertebrates assume dominance, and recovery of overexploited predators is impaired. Continued research aimed at integrating across trophic levels is needed to understand and forecast the ecosystem effects of changing oceanic predator abundances, the relative strength of top-down and bottom-up control, and interactions with intensifying anthropogenic stressors such as climate change.
