The goal of this paper is to evaluate commonly held criticisms of the practice of ability tracking in high school mathematics. To do so, I employ data from the National Education Longitudinal Study of 1988 and follow-ups to model classroom selection and education production. This paper will focus only on the causes and effects of tracking on students who were tracked as low-ability in eighth grade. From this, we can see how many students, if any, switched out of the low-ability track by tenth grade and how various switches have affected their test scores in mathematics. I find that students exercise mobility between ability-tracks as late as tenth grade and that ability-track placement is largely determined by test scores. In addition, I find evidence that there would be minimal, if any, test score improvement among low-ability students if they were all moved to a class of heterogeneous ability.