On-line homework is becoming a larger part of mathematics classrooms each year. Thus, ways to maximize the effectiveness of on-line homework for both students and teachers must be investigated. This study sought to provide one possible answer to this aim, by requiring students to achieve at least 50% for any on-line homework assignment in order to receive credit. Research shows that students respond well to reasonably set high expectations, and coupling this with one of the primary advantages of on-line homework, the ability to rework assignments, provided the basis for this study. Data for this experimental study was collected from the spring semester of 2008 until the fall semester of 2009, and included student exam scores, the number of on-line assignments above and below the 50% threshold, and the number of times students accessed help features of the on-line homework system when given the ability to do so. Analysis at both the whole-class and cluster levels attempted to discern the effectiveness of the intervention. Results indicated that significantly fewer students settled for on-line homework scores less than 50% in the experimental semesters where the 50% requirement was in place than in the control semesters in which the requirement was absent. Certain clusters of students seemed to benefit even more than others from this higher expectation, leading to the possibility of differentiated instruction or differentiated interventions in the future. In addition to fewer sub-par on-line homework scores, students also demonstrated other positive traits, such as accessing the on-line help links more within the experimental semesters.