This paper reports the results of a pilot intervention designed to improve students' mathematics proficiency through self-efficacy training. Seventeen pre-first year college students participated in a five-week summer program that included whole class instruction, small group tutoring, and individual meetings with instructional coordinators. As part of the intervention, the students made self-efficacy judgments on each of ten daily quizzes and compared these judgments to their math quiz scores. In the individual meetings, the students identified short term goals, created and maintained self monitoring forms, and were introduced to a math heuristic: the math card. The data from the self-efficacy training intervention were then analyzed using a hierarchical linear model approach. Over time, students' achievement scores on a math proficiency exam improved significantly, as did their confidence levels about passing this exam. Students who participated in the self-efficacy intervention group outperformed students who were involved in the regular remedial classes. (Author)