Following a discussion of various methods of evaluating the effectiveness of developmental programs, this paper presents a research design for the evaluation of a developmental math program. First, the paper examines the objectives and benefits of conducting formative and summative program evaluations. The paper then identifies the single group pre-test/post-test comparison as the most commonly used method of evaluating remedial programs, and identifies the biases to which the method is vulnerable (e.g., test administration, student attitude, instrument, and external learning biases). Next, methods of reducing the influence of some of these biases are suggested. The paper then focuses on the marginally remedial/marginally exempted comparison, the remediated/exempted comparison, norm-group comparison, cross-program comparison, and historical comparison methods. The paper then recommends that program evaluators be objective, informed, comprehensive, pragmatic, political, selective, and prepared to compromise. The last sections describe the methods of data collection and analysis, and the findings of the developmental math program evaluation. The project (1) investigated the number of students who completed the math courses, the number who took a more advanced math course after completing the developmental course, and marks in advanced math courses; and (2) compared results from classes in which math anxiety was recognized and treated with those in which it was not. (LAL)