With a sample of 7,752 fourth- to seventh-grade students in 25 schools which were randomized at the school level to condition, this article reports experimental impacts of an enhanced version of Word Generation on student outcomes at the end of Year 1 and of Year 2. Word Generation employs analysis, synthesis, critique, and problem-solving activities to build students' academic language, perspective taking, and ultimately their reading comprehension. Results indicate that the program improves the proximate outcome of academic vocabulary included in the curriculum after the first and second year of implementation for all students (Y1 effect size (ES) = 0.22, Y2 = 0.28 for fourth and fifth graders; Y1 ES = 0.13, Y2 = 0.16 for sixth and seventh graders). For those in the fourth and fifth grades, their perspective positioning skills also improved at the end of the first year (ES = 0.14), and their academic language skills (ES = 0.06), perspective articulation and positioning skills (ES = 0.12, 0.19), and reading comprehension (ES = 0.15) improved at the end of the second year. Among sixth and seventh graders, there were improvements in perspective positioning skills (ES = 0.19) and reading comprehension (ES = 0.10) at the end of Year 2. Effects after both Years 1 and 2 were stronger in high-exposure classrooms.