We used a combination of fleld and laboratory techniques to examine the relative importance of food webs based on marsh detritus, benthc algae, or phytoplankton in supporting growth of the blue crab Callinectes sapidus. We conducted a laboratory experiment to compare the growth of newly metamorphosed juveniles fed natural diets from potential settlement habitats such as marshes. The experimental diets consisted of zooplankton, Uca pugnax and Littoraria irrorata tissue, a mixture of plant detritus and associated meiofauna and detritus only. Crabs fed the zooplankton diet showed the fastest growth and reached a mean dry weight of 32.4 mg, from an initial dry weight of 0.8 mg, during a 3 wk period. Based on the isotopic composition, juvenile crabs obtain carbon and nitrogen from various food sources. For example, crabs fed zooplankton obtained their nutrition from phytoplanktonderived organic matter, consistent w t h zooplankton feedng on phytoplankton. The mean S13C values for juveniles fed detritus and detritus-plus-meiofauna were considerably lighter (613C = -19%0), than that of their respective diets (6'3C = -16%), suggesting that crabs were selectively ingesting prey items that obtain their nutrition from an isotopically lighter carbon source like phytoplankton. Conversely, crabs fed U. pugnaxor L. irrorata had isotopic ratios (6I3C = -16 to -14%) consistent with these species feeding on isotopically heavier marsh grass carbon. Isotopic ratios of crabs collected in the field appeared to corroborate the experiment and suggest that either Spartina alterniflora detritus or benthic algae-based food webs supported juvenile crab growth in marsh environments, whereas phytoplankton-based food webs dominate habitats more closely associated with the main estuary.