We tested whether predation, disturbance and food resources controlled the density and taxonomic composition of benthic macrofauna in salt marsh tidal creeks by manipulating predator and food abundance. Cages that excluded fish, crabs, and shrimp were placed in two urea fertilized and two control creeks at Great Sippewissett Marsh, MA, in late May 1979 and were sampled over a 15-week period.Densities of macrofauna peaked in early summer, then decreased markedly in natural sediments and inside partial cages that provided access to predator/disturbers. Inside predator exclusion cages, macrofaunal densities increased and remained high throughout the experiment. Comparisons of densities in natural and caged sediments indicated that predation and/or disturbance by fish, crabs, and shrimp depleted populations of Paranais litoralis, Streblospio benedicti, Hydrobia sp., Nematostella vectensis, and Capitella spp. during summer. In late summer few individuals remained in the natural community and no species was common. Bulk sediment properties were not significantly changed by cages.The urea fertilizer treatment increased chlorophyll a concentration and percent nitrogen and carbon in sediments. Despite improved food value of fertilized sediments to deposit feeders, no increases in density or biomass of macrofauna were detected in fertilized creeks relative to controls. This indicates that food does not limit macrofaunal abundance in this habitat, even in the absence of predator/disturbers. Fertilization did prompt changes in the species composition of the macrofauna indicative of a stressed community. The oligochaete P. litoralis was the numerical dominant in fertilized creeks, which averaged fewer species per sample than control creeks. A polychaete, S. benedicti, predominated in control creeks.