Educational games have the potential to provide motivating, effective training; however, the efficacy of these systems is unclear, and evaluations often fail to identify the relative impact of individual differences on learning outcomes. The current study aims to address these issues by comparing the learning gains from an educational game (iSTART-ME) and an intelligent tutoring system (iSTART). High-school students (n = 125) received comprehension strategy training from the two systems, and results indicated that both training environments yielded significantly better scores on posttest performance and learning measures than students assigned to a time-delayed control condition. Additionally, for both training conditions, students with a low prior 'commitment to reading' exhibited the highest performance improvements. Overall, results indicate that educational games can produce learning equivalent to intelligent tutoring systems, and that this training can provide a means to overcome initial deficits for students with a low 'commitment to reading'.