Pupil activity was observed in three Title I projects to determine the frequency of certain essential activity categories, to compare those frequencies with the supervisors' ideal, to assess relationships between pupil activity and achievement, and between implementation and achievement, and to compare supervisor ratings of implementation with observed implementation. Results indicated that pupil activity frequencies were different among the 13 to 37 sites within projects, that achievement varied significantly between sites, that there was no consistent relationship between pupil activity variables and achievement, and that neither measure of project implementation was related to achievement. Recommendations for future research are made. (Author)