Flux declines versus time during cross-flow microfiltration of wheat starch suspensions were modeled using fuzzy logic systems under different conditions of transmembrane pressure, flow rate, and suspension concentration. Fuzzy modeling is one of the most efficient methods for empirical modeling, especially for non-linear systems, and does not require any prior knowledge about the process. The influence of the process parameters on the dynamic behavior of the permeate flux was determined by comparing the permeate flux decline versus the volume curves, considering a constant value of the volumetric concentration factor during batch cross-flow filtration. At the transient stages of the filtration process, the permeate flux was shown to be strongly dependent on all process parameters, within the examined range.