This report considers the effectiveness of the Follow Through Program as manifested in performance on the Philadelphia city-wide tests, spring 1972. Kindergarten students received the Stanford Early School Achievement Test Level I; First and second grade students received the Metropolitan Achievement Test; Primary II respectively; and third grade students received the Iowa Tests of Basic Skills, Level A. The analysis in Part I, the cross-sectional view, revealed that the total program seems to be most effective in the first grade areas of total mathematics and reading. District groupings of Follow Through schools often show higher performance than their respective districts. District I groupings (equivalent to the Philadelphia Process Model) and District 4 groupings (solely Behavior Analysis Model schools but only two of the three schools in this model) appear to show this contrast more often than the other district groupings. However, these districts' groupings are homogeneous model-wise, in contrast to the District 2 and 5 groupings. The quasi-longitudinal view produced results in a pattern similar to the cross-sectional view, but demonstrates still higher performance, and seems to provide some initial evidence for a positive continuance-achievement relationship in the program. (Author/JM)