This paper presents a discussion of evaluation methods which can be used in preschool programs and emphasizes the importance of considering both outcome and process variables. Emphasizing the limitations of research which utilizes only outcome measures, the paper focuses on the benefits of an analysis of ongoing processes and interactions within early childhood programs. Two methods for observing and recording process variables are described. The first involves an open system consisting of actual descriptions or recordings of specimen behaviors of teachers and children. The second, a closed system, involves coded descriptions and includes time sampling, event sampling and trait ratings. The paper suggest that these methods of process evaluation can be used effectively to describe teaching practices, to train teachers, to monitor instructional programs and to investigate relationships between classroom activities and pupil growth. Methodological issues which arise in the use of process evaluation are also discussed. Included in this discussion are the problems of instrument choice, observer reliability, and the consideration of the degree of control the researcher has over the selection and supervision of the program being evaluated. It is suggested that only through both outcome and process evaluation will it be possible to develop the best procedures for educating young children in preschool programs. (BD)