Our purpose for this quasi-experimental study was to evaluate the short-term and maintenance effects of the self-regulated strategy development writing instructional model by Graham and Harris with 7th-grade students in an urban, ethnically diverse Title I middle school. We compared the writing skills of our intervention students with those of students in a control school. For 5 weeks, we coached teachers at the intervention school in a strategy for persuasive writing. Teachers at the control school also taught persuasive writing but used traditional instruction. We used a pre/posttest design to measure short-term growth, and we collected a maintenance writing score 2 months after the intervention. We used gain scores from pretest to posttest and from posttest to maintenance in a hierarchical linear modeling analysis. We found no differences between the 2 groups from pretest to posttest; however, scores between posttest and maintenance showed that the self-regulated strategy development students scored significantly higher than students in the control school.