The CROSMOR-model (Van Rijn, 1997, 1998), representing wave propagation, refraction, shoaling, breaking, associated sand transport and morphology in cross-shore direction based on a probabilistic approach, has been applied to simulate the hydrodynamics and the nearshore bar behaviour for three experiments in the large-scale Delta flume o f Delft Hydraulics. The CROSMOR model has been applied with updated submodels for hydrodynamics and sand transport. The update of the submodels is related to: • the description of the wave orbital velocity using the modified Isobe method; • the effect of bed roughness and wave breaking on sediment mixing (diffusivity); • the representation of the high frequency oscillatory suspended transport. The effects of the updated models on bed morphology will be evaluated. Conclusions • simulation of wave height is almost perfect for tests 1A and 1B; the shoaling wave height in test 1C is slightly too large; • simulation of the undertow shows reasonable results (within 30% error) for test 1A and 1B; • computed values in test 1C show good agreement with measured values in the trough zone, but computed values are too large (factor 2) seaward of the bar; • computed onshore-directed values are systematically too large (about 10% to 20%) in test 1A and 1B, especially just landward of the bar crest where the waves are broken; • computed peak offshore orbital velocity are in good agreement with measured values in test 1A and 1B, but substantially too large (30%) in the bar crest zone of test 1C; • in test 1B the computed transport rate based on variable bed roughness is onshore-directed just seaward of the bar and offshore-directed in the bar crest-trough zone and again onshore-directed just landward o f the trough zone; the computed transport rates are of the right magnitude, but the computed transport rate distribution at the bar crest is too peaked and somewhat shifted in seaward direction at the end of the test; the computed sand transport is less realistic if the effect of the oscillatory suspended transport component is neglected; • in test 1C the onshore-directed transport rates based on variable bed roughness are quite well represented seaward of the bar crest; at the bar crest the sand transport is much too small; the offshore-directed sand transport just landward of the bar crest is reasonably well represented; the computed sand transport is less realistic if the effect of the oscillatory suspended transport component is neglected; the computed sand transport can be improved substantially by increasing the local onshore-directed near-bed orbital velocities with 20% (local slope larger than 0.04) based on results of sensitivity computation; • the Bagnold sand transport formula yields onshore-directed transport rates at all locations and the computed transport rates differ significantly from measured values. • realistic simulation of bar development in all three tests requires the simulation of variable bed roughness along the profile; pronounced ripples are present in the trough of the bar and the effective roughness of these ripples should be taken into account by using a larger roughness value in the bar trough zone; • the effect of the oscillatory suspended transport on the short term bed evolution is not very substantial, because the transport gradients are not very much affected by this transport component; the effect on long term bed evolution will however be more significant because the absolute value of the incoming transport (at x=0) is modified by including the oscillatory suspended transport (additional onshore transport component); • the simulated bar development is almost perfect if the onshore-directed near-bed orbital velocity is increased with 20% at that part of the profile where the local slope is larger than 0.04 (1 to 25); this results in a local increase of the asymmetry-related bed and suspended transport; • application of the Bagnold sand transport formula does not result in sufficient bar development, because the offshore-directed sand transport component is not accurately modelled.