We use a regression-discontinuity design to study the effects of double blocking sixth-grade students in reading and mathematics on their achievement across three years of middle school. To identify the effect of the intervention, we use sharp cutoffs in the test scores used to assign students to double blocking. We find large, positive, and persistent effects of double blocking in reading, but, unlike previous research, we find no statistically significant effects of double blocking in mathematics either in the short run or medium run. The following is appended: Differences in Means of Control Variables around the Cutoff for Assignment to Double Block in Math.