The practice of tracking and ability grouping--the division of students into separate tracks, classes, and groups for instruction based on their purported interests and abilities--has long been debated. Evidence from decades of research indicates that tracking magnifies inequality between high and low achievers without raising achievement overall, as high achievers perform better in tracked systems while low achievers perform worse, compared to similar students in mixed-ability contexts. These findings have been sustained in recent work, which has also advanced in three areas. First, international studies have yielded results that are generally consistent with those previously found for the U.S. and U.K. Second, new attempts to reduce or eliminate tracking have suggested ways in which some of the obstacles to reducing the practice may be overcome. Third, new work on classroom assignment and instruction has identified approaches that may capture the benefits of differentiation for meeting students' varied needs without giving rise to the consequences for inequality that commonly accompany tracking and ability grouping. These findings in turn call for new research and experimentation in practice. (Contains 5 footnotes.) [This paper was prepared for the "The Routledge International Handbook of the Sociology of Education," edited by Michael W. Apple, Stephen J. Ball, and Luis Armand Gandin. New York: Routledge, in press.]