This study investigated the usefulness of behavior engagement as an outcome measure for discerning the efficacy of intervention efforts, broadly conceived as specific attempts at affecting child performance through environmental manipulation. Engagement refers to the amount of time children spend interacting with the environment in a developmentally appropriate manner. Levels of engagement were measured using the Planned Activity Check (PLA-Check) among preschool mentally retarded, handicapped, at-risk, and nonhandicapped children participating in two different types of programs. One was traditional in orientation and the other nontraditional. The traditional program focused on the children manifesting discrete forms of criterion behavior, while the nontraditional program emphasized engagement as the focus of classroom efforts. PLA-Checks were taken during four classroom activities on 2 separate days. The results showed the engagement levels differed as a function of program type and classroom activities within programs. Engagement levels were found to be consistently more stable in the nontraditional program. The findings were discussed in terms of the usefulness of engagement as a behavioral measure of the efficacy of early intervention.