Twenty-nine peer-reviewed journal articles that analyzed intervention implementation in schools using single-case experimental designs were meta-analyzed. These studies reported 171 separate data paths and provided 3,991 data points. The meta-analysis was accomplished by fitting data extracted from graphs in mixed linear growth models. This analytic approach was adopted to provide a model whose assumptions were met and could model both level and trend. The extent to which interventions were implemented as planned in baseline was low with a downward trend. Performance feedback, performance feedback augmented with additional procedures, and self-monitoring with environmental supports were found to be highly effective for improving intervention implementation. Follow-up meetings without review of implementation data were ineffective. Maintenance data were positive across studies. Implications of the results for future research and practice are discussed.