The paper examines several published mathematical models of electrostatic precipitator (ESP) performance--the Deutsch-Anderson, Matts-Ohnfeldt, Cooperman, Robinson, Southern Research, and Research Triangle Institute sectional models--and compares their predictions of ESP performance against measured data. The analysis shows the common areas of the models and the specific vagaries of each. In all current models, it is seen that unmodeled effects can dominate the accuracy of the modeling process, but that a 'calibration' of a given model can make the model useful for those effects it handles well. The values of important parameters for each model are tabulated so that the reader may use them as summary descriptions of the ESP data.