Some ninth-grade students from three Knoxville City high schools have continuously performed substantially less well than those from other city schools in the Tennessee State proficiency tests. A task force developed a plan to work with grade 8 in the middle schools and grade 9 in the high schools to prepare students to perform better on state proficiency and citywide basic skills tests. Pretest/posttest data from experimental (N=170) and control groups (N=175) were analyzed in an attempt to determine the impact of the project on pupil performance. Three groups (teachers, pupils, and parents) were to receive various "treatments" that included many elements of the effective schools efforts. Gross measures such as total percentage passing the language arts and mathematics proficiency tests in various high schools--and comparisons of these scores over time--were used as preliminary indicators of project progress. Additional data included anecdotal reports from teachers, logs and timelines of activities, comments of project researchers, and teacher questionnaire data. Major gains on the percentage of pupils passing the proficiency test were observed in 1985. An analysis of covariance supported the assertion that the sample of project students significantly outperformed a similar sample of control students. Tables and graphs illustrate the data. Appended are 12 references. (MLF)