The study employed a randomized, controlled trial design to examine the effect of an instructional approach to improve the inferential comprehension of students who were good and poor readers. The instruction consisted of activating background knowledge, answering questions, and making predictions, and the control group received the typical instruction in the classroom. The intervention lasted for ten weeks.Forty fourth-grade students of diverse socioeconomic backgrounds from an elementary school in a small town in Maine participated in the study. Half of the students were good readers and half were poor readers based on Stanford Achievement Test results and teacher judgment. The results showed that the instructional approach of activating background knowledge, answering questions, and making predictions improved students' comprehension. The intervention was particularly beneficial to poor readers, who outperformed their peers in the control group.