This paper discusses the experiences and insight derived from the 1985-1986 Statewide Evaluation of the Early Education Program of Handicapped Children in Louisiana, the multimethod evaluation of preschool intervention projects for handicapped and at-risk children. The program served almost 5,000 handicapped preschoolers and their parents, providing 17 different kinds of services in home-based, center-based and combination program formats. The multiple-model design applied in this program evaluation consisted of four segments: (1) a prefatory naturalistic inquiry to provide a foundation for the design matrix and to check the appropriateness of the evaluation components; (2) first-wave case studies and a qualitative approach to develop data-collection strategies and instruments that were specific to the identified service roles; (3) second-wave case studies or a combination of qualitative interviews and surveys/questionnaires to complete the final drafts of the 13 role-specific instruments; and (4) third-wave research or the integration of the case study/observation component with a major data collection effort that consisted of service provider and parent surveys, an interagency focus group interview, and the aggregation statewide of class and individual student profiles. The use of this multimethod approach facilitated the identification and investigation of 14 program-related issues. It also resulted in a high level of cooperation on the part of state, regional and local program personnel, parents, and kindergarten teachers as evidenced by high return rates on surveys and questionnaires and by requests for additional feedback. (JAZ)