Polymerization in the presence of templates, followed by their consequent removal, leads to structures with cavities capable of molecular recognition. This molecular imprinting technology has been employed to create porous polymers with tailored selectivity for adsorption, chromatographic separations, sensing, and other applications. Performance of these materials crucially depends on the availability of highly selective binding sites. This parameter is a function of a large number of processing conditions and is difficult to control. Furthermore, the nature of molecular recognition processes in these materials is poorly understood to allow a more systematic design. In this work we propose a simple model of molecularly imprinted polymers mimicking the actual process of their formation. We demonstrate that a range of molecular recognition effects emerge in this model and that they are consistent with the experimental observations. The model also provides a wealth of information on how binding sites form and function in the imprinted structures. It demonstrates the capability to assess the role of various processing conditions in the final properties of imprinted materials, and therefore it can be used to provide some qualitative insights on the optimal values of processing parameters.