The study of how domestic social service programs have been implemented remains an embryonic field in search of future research directions. Relying largely on their exploratory analysis of the implementation of Title I of the Elementary and Secondary Education Act from a 13-year perspective, the authors argue that a longitudinal approach synthesizing evidence from a confluence of sources merits serious consideration as a research approach for studying implementation in the 1980s. Their longitudinal investigation of two major program requirements for Title I suggests that shifts in administrative policies and procedures in a decentralized policy context are typically dominated by incremental changes; that when aggregated across time, the incremental changes have resulted in substantive structural modifications in implementation policies and practices; and that the pace and direction of these changes have been most affected by a shifting interest group scenario, feedback from previous implementation stages, and broad social movements. The essay also outlines procedural considerations for conducting longitudinal implementation investigations and makes explicit some of the limitations and problems inherent in such an approach. (Author/IRT)