Although considerable research examines the impact of grade retention on student academic achievement, various shortcomings exhibited across these studies make it difficult to derive any concrete generalizations about the effectiveness of requiring students to repeat a grade. By utilizing data based on a cohort of all low-achieving elementary students in the state of Texas over a number of years, one can overcome some of the weaknesses observed in previous studies on grade retention and social promotion. The purpose of this study is to ascertain whether holding low-achieving students back a year in grade contributes to enhancing academic performance. This analysis is based upon archival data supplied by the Texas Education Agency, consisting of anonymous student-level Texas Assessment of Academic Skills (TAAS) records of every student in the tested grades between 1994 and 1999. The authors sought to determine the academic effects of retention in grade and social promotion on subsequent school performances, following initial failure on the Texas mandatory standardized reading test. The overall findings suggest that retaining low-performing third-grade students an additional year in grade was not harmful to their later academic performance. Analyses that attempted to control for initial differences between retained and socially promoted pupils revealed that retention appeared to help low-achieving students pass the state reading examination more quickly than students who were socially promoted. The authors conclude by outlining some conditions that they argue are essential for effective policy, effective interventions, and disaggregated accountability systems. Comments by Andrew Rotherham and Lorrie A. Shepard are presented. (Contains 1 figure, 8 tables, 52 notes.)