Abstract: Receptor sites capable of distinguishing between molecules having minor structural differences can be prepared by noncovalent imprinting of templates in network polymers. Initially, these polymers are prepared considering functional group complimentarity between a functional monomer and the template. The performance of these "first generation" materials is often unacceptable and a careful optimization of the variables in the imprinting process is therefore needed in order to reach the desired level of affinity and selectivity for the target compound. Depending on the separation requirements, the imprinted materials arc further associated with problems. i.e. non-linear adsorption isotherms slow mass transfer kinetics, low sample load capacities, that may need to he solved. In the above context, this talk will address options available at each stage in the process of preparing and evaluating an imprinted polymer. Recent examples of how affinity may be enhanced by applying pressure during polymerization and how the hydrophobic effect may contribute, either specifically or nonspecifically, to the observed binding will be given