We compared spatial variation in the abundance of I n g George whiting Sillaginodes punctata post-larvae in the southern part of Port Phillip Bay, Australia, with predictions of hydrodynamic and dispersal numerical models that included passive transport, and vertical migratory behaviour previously observed in the field. Post-larvae were sampled at 20 sites on cruises in the spring of 1994 and 1995. Modelling included a 'passive' case with particles mixed through the water column, 'active' behaviour mirroring die1 and tidal variation in the field, and 'day/night' behaviour where behaviour in the field was averaged across tides. Correlations between model simulations and post-larval distribution were highest in 1995 and were similar amongst the 3 simulations (58% of variation explained for both the passive and dayhight cases). In 1994, variation in the spatial distribution explained was highest for the passive case, intermediate for the active case and lowest for the day/night case. Prevailing winds were quite different between years, with westerlies prevailing in 1994 but significant periods of easterly winds in 1995. These differences were reflected by particle cllstributions from simulation~ including behaviour, but were not reflected in post-larval distributions. A negative correlation was found between post-larval abundances and distance from shore. When model predictions and distance from shore were combined in a multiple regression, approximately 70% of the spatial variation in post-larval abundance was explained in 1995. The results imply that although the passive transport model was an excellent predictor of post-larval abundance in both years, observed vertical migration was not influencing transport, and post-larvae were closer to shore than expected, possibly due to behaviours other than vertical migration. The close association of S. punctata post-larvae with the coastline provided a mechanism for transport further into the bay, against the prevailing wind field.