Food habits in Pamlico Sound, North Carolina, are poorly described despite the estuary's large size and importance as nursery and fisheries habitat. We conducted the first multi‐year, multispecies food habits study in Pamlico Sound, sampling the stomach contents of 16,913 predators representing 25 species. Predators were sampled from fisheries‐independent trawl and gill‐net surveys. We used multivariate analyses to compare diets between surveys, used agglomerative hierarchical cluster analyses and similarity profiles to identify significant trophic guilds, and identified forage fish using multiple approaches (qualitative classification criteria, connectance, and supportive role to fishery ecosystems [SURF]). The diets of predators sampled from the trawl survey were significantly different than predators sampled from the gill‐net survey. Mysids and anchovies were more important for trawl‐caught predators, with the majority of those predators belonging to nonpiscivorous guilds. Half of the gill‐net survey predators were piscivorous and relied more heavily on Atlantic Menhaden Brevoortia tyrannus and sciaenids. Differences in the level of piscivory between the surveys are most likely a result of larger predators being sampled in gill nets relative to trawls. There was little agreement among approaches in forage species identification, and only anchovies in the trawl survey were identified as a forage species using all approaches. Quantitative metrics identified forage species (e.g., Spot Leiostomus xanthurus and invertebrates) that were not identified by qualitative classification criteria. Our work shows the effect of gear size selection on estimates of predator diets and the need to use a variety of gears that sample a wide range of predator sizes. Additionally, the identification of forage species requires an evaluation of criteria outside of life history characteristics and a greater emphasis on the contribution of a prey species to a predator's diet.