Why are some students more likely to succeed in or fail mathematics than their peers, in spite of similar learning opportunities? Frequently such differences have been framed within liberal or even conservative discourses that identify purportedly innate dispositions, such as the ability or intelligence, as the root of differential success. However, a significant body of sociological research challenges such fundamental assumptions by demonstrating that success or failure is not random, but rather, closely linked to the background (gender, social class, language, or culture) of the student (Boaler, 1997c; Cooper and Dunne, 1999; Lubienski, 2000; Zevenbergen and Lerman, 2001). This article poses a challenge to the opportunities offered by practices that group students according to perceived ability. Using the theoretical constructs of French sociologist Pierre Bourdieu, the author argues that inserting students into particular ability groups creates learning environments that influence how students come to see themselves as learners of mathematics, i.e., the construction of mathematics identity, which can have implications for future learning. A total of 96 students, from six Australian schools, were interviewed using a semi-structured interview schedule. The students were in years 9 and 10 in secondary school, so were generally in the age range of 14-16 years. Students self-reported their levels of "ability" determining their location in particular class groupings. All schools had practices whereby students were grouped according to ability. Part of the study asked students to describe a typical mathematics lesson. Further questions centered on their experiences in their classrooms. Overall, the results indicated two key practices that form the basis of the students' experiences in secondary school mathematics teaching. The first practice, that of the practice of teaching, was unanimously of a particular form. The second practice, that of ability grouping, was evident in all schools. The experiences of the students fell clearly into distinct categories, whereby the students in higher groups felt that they were blessed with high-quality experiences, while students in lower groups reported that their experiences were quite negative.