A hybrid receptor model is a specified mathematical procedure which uses not only the ambient species concentration measurements that form the input data for a pure receptor model, but in addition source emission rates or atmospheric dispersion or transformation information characteristic of dispersion models. By utilizing more information hybrid receptor modeling promises improved source apportionment estimates or, more fundamentally, consideration of problems that are inaccessible in terms of classical receptor modeling. Two examples of hybrid receptor models are reviewed to illustrate the variety in possible approaches. Some hybrid receptor modeling results are given for the comprehensive ambient-source meteorological data base collected at Deep Creek, Maryland, during summer 1983. By using source and ambient measurements of selenium about 33% of the sulfate impact at the receptor site during a day-time period can be associated with emissions from a coal-fired power plant 50 KM upwind.