We used measurements of larval abundance and longitudinal, tidal veloc
ity to calculate flux of weakfish Cynoscion regalis larvae on the spaw
ning grounds in southern Delaware Bay, USA. Mean flux was significantl
y greater during flood phase for both early-stage and late-stage larva
e, but there was no difference between flood and ebb flux for yolksac
larvae. We developed a new concept, larval velocity, which quantifies
the subtidal velocity with which larvae are transported in an estuary.
We also constructed a mathematical model that allows comparison of me
an larval velocity and mean subtidal current velocity in an estuary. L
arval velocity and water velocity were similar at all depths in Delawa
re Bay for yolksac larvae. Larval velocity was always landward (up-est
uary) for more advanced larval stages and was considerably greater tha
n subtidal water velocity, especially near the surface. Results are ex
plained in terms of selective tidal stream transport.