We used carbon stable isotopes to examine the influence of the size of saltmarsh patches on the trophic contribution of saltmarsh grass and mangroves to 2 species of resident crabs (Parasesarma erythrodactyla and Australoplax tridentata). Crabs were collected at different distances across the saltmarsh-mangrove interface at each of 10 saltmarshes of different sizes (0.01 to 8.10 ha) adjacent to mangrove forests (each >4 ha). The d13C values of crabs at all 10 marshes fitted a sigmoidal curve, with rapidly changing d13C values across the saltmarsh-mangrove interface (the transition zone). The size of saltmarsh patches had a significant effect on d13C values of P. erythrodactyla collected in the saltmarsh, with a similar trend shown by A. tridentata. On large saltmarshes (>0.4 ha), d13C values of crabs (P. erythrodactyla, -15.9黠A. tridentata, -15.4) collected >5 m onto the saltmarsh matched that of the saltmarsh grass Sporobolus virginicus (-15.5). Carbon movement and assimilation by crabs was limited to <5 m. On small saltmarshes (<0.3 ha), d13C values of saltmarsh crabs (-18.1 and -16.8, respectively) were depleted, indicating assimilation of carbon from S. virginicus, but also from a more depleted allochthonous source, e.g. mangroves (-28.1), microphytobenthos (-23.7), or phytoplankton (ca. -20). Saltmarsh patch size did not affect the extent of carbon movement or assimilation by crabs in mangroves. Given that habitat patch size can influence pathways of carbon supply to invertebrates, the role of estuarine habitats in food webs cannot be assumed to be independent of the size and configuration of habitats.