A quasi-experimental study of a novel, cross-content area vocabulary intervention program called "Word Generation" showed significantly greater growth among 6th- to 8th-grade students in schools implementing the program than in comparison schools, on a curriculum-specific test. Furthermore, the language-minority students in the treatment, but not the comparison, schools showed greater growth than the English-only students. Improvement on the curriculum-specific test predicted performance on the Massachusetts Comprehensive Assessment System (MCAS) English language arts assessment, again only for students in the treatment schools. Recognizing the need to implement a more rigorous experimental study of this program, nonetheless we conclude that participation in the intervention, with its focus on deep reading, comprehension of current-events topics, productive classroom discussion, developing arguments, and producing persuasive essays, was a plausible contributor to student performance on the MCAS. (Contains 2 figures and 11 tables.)