The New Zealand cockle Austrovenus stutchburyi, whose foot is commonly infected by the digenean trematode Curtuteria australis (Echinostomatidae), is often found heavily infected and unable to burrow on the sediment surface of tidal flats. This has been interpreted as a Curtuteria-manipulation with the purpose of increasing the transmission of the parasite to shorebirds acting as final hosts. Using a field-experimental approach the alternative hypothesis was tested, that surface-dwelling cockles, caught on the surface for other reasons than parasites, accumulate larval C. australis at a higher rate than buried cockles. During the 3-month experiment, larval trematodes accumulated with a rate of approximately 0·5 metacercariae/day in both surface and buried cockles. The result strengthens the manipulation hypothesis indirectly by rejecting the alternative hypothesis. The metacercariae were unevenly distributed along the cockle-foot, with about 4 times as many cysts being found in the tip than in either the mid or hind part of the foot. In light of existing knowledge of the burrowing behaviour and apparatus in bivalves, and a negative relationship between foot mobility and infection intensity, it is suggested that C. australis manipulates its host through a mechanical obstruction of foot muscles and the dynamic hydrostatic skeleton, both necessary for successful burrowing.