Here we present a modified version of the exploratory numerical model originally presented by Murray and Thieler (2004) and use it to further investigate the development of sorted bed forms on the inner continental shelf. The new version of the model is based on widely accepted parameterizations for hydrodynamics and sediment transport that are more detailed and empirically tested than the parameterizations used in the original model. The shallow water wave equations are replaced by the general equations, and changes in water depth related to morphological development are accounted for. This model reproduces the main results of the previous one, supporting the hypothesis that sorted bed forms are a self-organized pattern and develop as a result of a coupling between bed composition and sediment flux. The sensitivity of the model predictions, evaluated in terms of the geometry of the sorted bed forms and their growth rate, with respect to external forcing variables and internal model parameterizations, shows increasing (decreasing) sorted bed form spacing (height) with increasing wave height and period. The opposite occurs for increasing water depths. More complicated behavior is observed when the ratio fine-to-coarse sediment diameter (or the percentage of fine and coarse material) is varied. The shape of the suspended sediment and current profiles and the formulation used to predict the geometry of wave-generated ripples can significantly affect (or even inhibit) growth rates of sorted bed forms. Results from a sensitivity analysis show that there are deficiencies in the description of small-scale processes that limit the ability to quantitatively predict sorted bed form evolution, highlighting gaps in our knowledge of sediment transport over mixed grain beds that need to be filled. Copyright 2007 by the American Geophysical Union.