The relative influence of consumers (top down) and resources (bottom up) on the distribution and abundance of organisms remains a key question in ecology. We examined the relationships between consumer and resource variables along a productivity gradient for a dominant predator-prey interaction in a marine soft-sediment system. We 1) quantified density and size of the clam Macoma balthica (prey species) in six replicate sites at each of four habitat types (shallow mud, deep mud, muddy sand and detrital mud) in the Rhode River, Chesapeake Bay. We selected one habitat type of high food availability and clam density (shallow mud) and another of low food availability and clam density (muddy sand) for manipulative experiments. Then, we 2) measured M. balthica survival and growth through transplants, 3) measured food availability as sedimentary organic carbon content, 4) quantified predator density, and 5) calculated predator foraging efficiency in the two habitat types. Clam density in the four habitat types differed and was related to sedimentary carbon availability and predator density. One of the habitats, detrital mud, appeared to be a population sink because it only held juvenile Macoma that never survived to reproductive age. Macoma size and growth, and predator (mainly blue crab Callinectes sapidus) densities were positively correlated with productivity and were higher in shallow mud than muddy sand. In contrast, Macoma mortality, local 'interaction strength', and predator foraging efficiency were lower in the productive habitat (shallow mud). Thus, predation intensity was inversely correlated with productivity (food availability); consumer and resource effects differed by habitat type; and, at a relatively small spatial scale, consumer and resource forces jointly determined population dynamics in this soft-sediment marine system.