Tidal marsh wetlands provide important foraging habitat for a variety of estuarine fishes. Prey organisms include benthic–epibenthic macroinvertebrates, neustonic arthropods, and zooplankton. Little is known about the abundance and distribution of interior marsh macroinvertebrate communities in the San Francisco Estuary (estuary). We describe seasonal, regional, and site variation in the composition and abundance of neuston and benthic–epibenthic macroinvertebrates that inhabit tidal marsh channels, and relate these patterns to environmental conditions. We also describe spatial and temporal variation in diets of marsh-associated inland silverside, yellowfin goby, and western mosquitofish. Fish and invertebrates were sampled quarterly from October 2003 to June 2005 at six marsh sites located in three river systems of the northern estuary: Petaluma River, Napa River, and  the west Delta. Benthic/epibenthic macroinvertebrates and neuston responded to environmental variables related to seasonal changes (i.e., temperature, salinity), as well as those related to marsh structure (i.e., vegetation, channel edge). The greatest variation in abundance occurred seasonally for neuston and spatially for benthic–epibenthic organisms, suggesting that each community responds to different environmental drivers. Benthic/epibenthic invertebrate abundance and diversity was lowest in the west Delta, and increased with increasing salinity. Insect abundance increased during the spring and summer, while Collembolan (springtail) abundance increased during the winter. Benthic/epibenthic macroinvertebrates dominated fish diets, supplemented by insects, with zooplankton playing a minor role. Diet compositions of the three fish species overlapped considerably, with strong selection indicated for epibenthic crustaceans—a surprising result given the typical classification of Menidia beryllina as a planktivore, Acanthogobius flavimanus as a benthic predator, and Gambusia affinis as a larvivorous surface-feeder. Fish diets were influenced by position along the estuarine gradient and season. Overall, our data show that local-scale site effects and marsh position within the estuary influence invertebrate community composition and abundance. Additionally, we show that restoring marsh ecosystems can subsidize fishes similarly to reference marshes. We, thus, recommend that managers focus on the ability of restoring marshes to produce food subsidies for target species when planning and designing tidal marsh restoration projects, especially those targeted for food web support.