| Although most people agree that in order to teach mathematics effectively, teachers must understand mathematics well themselves, past effort to show the relationship between mathematics teachers' subject matter preparation and their students' performance on mathematics tasks has largely been unsuccessful.;This study focused on the interaction effect between teachers' mathematics preparation and the thinking level of the mathematics tasks on student performance. The major hypothesis of this study posits that the effect of 7th grade mathematics teachers' subject matter preparation on student performance depends on the thinking level of the mathematics tasks, and that as the thinking level of the mathematics tasks increases, differences in student test performance become in favor, or more in favor, of more specialized mathematics teachers.;The sample of this study consisted of 7th grade students from 33 pairs of matched schools. One school in each pair mostly used secondary-prepared mathematics teachers to teach 7th grade mathematics while the other school mostly used elementary-prepared and/or out-of-field teachers. Average student responses to the Stanford-8 mathematics items, which were classified into low, medium, and high thinking levels, were used to analyze the effect of teachers' mathematics preparation.;Results from the statistical analysis comparing the difference between two trends of student performance supported the interaction hypothesis of this study. The interaction involved almost equal performance of students taught by secondary-prepared and elementary/out-of-field teachers on lower-level thinking mathematics problems but significantly better performance of students taught by secondary-prepared teachers on high-level thinking mathematics problems. The effect size of teacher mathematics preparation at the teaching of high-level thinking skills was found to be 0.49, which by convention is considered medium-sized. |