Abstract: A new and fast technique for screening combinatorial libraries of molecularly imprinted polymers is presented. The procedure is based on commercially available membrane modules which are rinsed with pre-polymerization imprinting mixtures. After the in situ polymerization and generation of MIP films on the PTFE membranes within the modules, the membranes are evaluated in terms of affinity towards the target molecule (template) R-(-)- phenylbutyric acid. Therefore, after template extraction from the freshly produced membranes a solution of this target molecule is flushed through the module. By analyzing the remaining analyte concentration in the permeate, the amount of analyte adsorbed on the membrane can be calculated and related to specific interactions with the molecular imprints. By this means, optimized recipes in terms of cross-linker to template ratios could be obtained in combination with the optimal porogen, when screening p-divinylbenzene or ethylene glycol dimethacrylate as cross-linker and porogens like acetonitrile, dimethylsulfoxide and methanol. (C) 2004 Elsevier B.V. All rights reserved