This article reports on a large scale randomized controlled trial to study the efficacy of a web-based intelligent tutoring system for the structure strategy designed to improve content area reading comprehension. The research was conducted with 128 fifth-grade classrooms within 12 school districts in rural and suburban settings. Classrooms within each school were randomly assigned to intervention or control groups. The intervention group used the intelligent web-based tutoring system for the structure strategy (ITSS) for 30 to 45 min each week as a partial substitute for the language arts curriculum for the entire school year. The structure strategy teaches students how to read and comprehend expository texts by identifying the text structure and creating strategic mental representations of the text. The web-based tutoring system delivered the structure strategy training with modeling, practice tasks, assessment, and feedback. The control classrooms used the school's language arts curriculum for the full language-arts time. Results show that the ITSS delivered structure strategy training improved reading comprehension measured by a standardized test and researcher designed measures.