This brief experimental study investigated the initial promise of an intervention designed to promote inference generation in adolescents with reading comprehension difficulties. The intervention, provided for nine sessions, included multisyllable word study, teacher explanation and modeling of inference generation and other comprehension processes, and having students practice by thinking aloud about text. Research questions addressed proximal effects on measures of the intermediate goals of the intervention and effects on reading comprehension. Participants were 48 ninth grade students with reading comprehension difficulties, randomly assigned to experimental or time-on-task control conditions. No significant group differences were detected; however, effect sizes in the moderate range indicated meaningful effects on some proximal measures, suggesting that further study of the approach is warranted.