The Knoxville City Schools (KCS) Proficiency Project ran from fall 1984 to spring 1987 in three Knoxville, Tennessee, middle schools and three high schools. Students in three of the eight KCS high schools were scoring consistently lower than pupils in the other schools on the grade 9 State Proficiency Test, especially in math (M) and language arts (LA). The primary intervention was a group of three proficiency teachers; each worked with one middle school (especially grade 8) and one high school (especially grade 9) to help teachers prepare grade 9 pupils in the three target schools to pass the M and LA portions of the test. This paper compares annual test results in the following groups of schools: (1) Proficiency Project, or target; (2) other schools; and (3) city, i.e., total. In 1982 there were great differences among the three groups of schools. After one year of the project in 1985, differences among the schools were small; the 1986 differences were a little greater than in 1985, but in 1987 the differences became comparatively less. The city scores for 1987 are greater than the 1982-84 average. The following explanations may account for this increase: (1) the target schools' scores pushed up the city scores; (2) the other schools' scores pulled up the city scores; or (3) a combination of the above. Data are presented on seven tables and several figures. A list of references is included as well as a list of prior reports of this project. Appendices include t-test analyses comparing pretest and posttest scores of selected groupings of respondents, and present the results of a questionnaire of math and language arts teachers' needs and concerns. (BJV)