In theory, carbon is highly mobile in aquatic systems. Recent evidence from carbon stable isotopes of crabs (Parasesarma erythrodactyla and Australoplax tridentata), however, shows that in subtropical Australian waters, measurable carbon movement between adjacent mangrove and saltmarsh habitats is limited to no more than a few metres. We tested whether the pattern in crab delta13C values across mangrove and saltmarsh habitats was explained by crab movement, or the movement of particulate organic matter. We estimated crab movement in a mark-recapture program using an array of pitfall traps on 13 transects (a total of 65 traps) covering an area of 600 m2 across the interface of these two habitats. Over a 19-day period, the majority of crabs (91% for P. erythrodactyla, 93% for A. tridentata) moved <2 m from the place of initial capture. Crab movement cannot, therefore, explain the patterns in delta13C values of crabs. delta13C values of detritus collected at 2-m intervals across the same habitat interface fitted a sigmoidal curve of a similar form to that fitting the delta13C values of crabs. delta13C values of detritus were 2-4 per thousand more depleted in saltmarsh (-18.5+/-0.6 per thousand), and 4-7 per thousand more depleted in mangroves (-25.9+/-0.1 per thousand) than delta13C values of crabs recorded previously in each habitat. Assimilation by crabs of very small detrital fragments or microphytobenthos, more enriched in 13C, may explain the disparity in delta13C values. Nevertheless, the pattern in delta13C values of detritus suggests that crabs obtain their carbon from up to several metres away, but without themselves foraging more then a metre or so from their burrow. Such detailed measurements of carbon movement in estuaries provide a spatially explicit understanding of the functioning of food webs in saltmarsh and mangrove habitats.