A variety of reading process models have been proposed and evaluated in reading research. Traditional approaches to model evaluation specify the workings of a system in a simplified fashion to enable organized, systematic study of the system's components. Following are several statistical methods of model evaluation: (1) empirical research on components, in which the researcher predicts behavior based on an analysis of the model's components; (2) tally sheet summarization of studies addressing particular issues; (3) meta-analytic techniques, a more sophisticated inter-study summarization than the tally sheet approach; and (4) causal modeling, in which researchers go beyond the constraints of correlation analysis to make statistical inferences concerning causation. Comparisons of reading models are often descriptive in nature rather than analytical. Leading texts on reading are generally comprised of key theoretical articles that provide readers with important and original introductory materials on various reading models. English/language arts education has long emphasized philosophical reading theories based on what reading should be, although several attempts have been made to evaluate reading models by constructing classification systems as well. The ultimate test of any reading model rests on whether it can be effectively applied in the classroom. A four-page reference list is provided. (JD)