Molecularly imprinted polymers (MIPs) are polymers that can be tailored with affinity and selectivity for a molecule of interest. Offsetting the low cost and ease of preparation of MIPs is the presence of binding sites that vary widely in affinity and selectivity. Presented is a review of methods that take into account binding site heterogeneity when calculating the binding properties of MIPs. These include the bi-Langmuir, Freundlich, and Langmuir-Freundlich binding models. These methods yield a measure of heterogeneity in the form of binding site affinity distributions and the heterogeneity index. Recent developments have made these methods surprisingly easy to use while also yielding more accurate measures of the binding properties of MIPs. These have allowed for easier comparison and optimization of MIPs. Heterogeneous binding models have also led to a better understanding of the imprinting process and of the advantages and limitations of MIPs in chromatographic and sensor applications.