According to Bourdieu, the culture of the most powerful classes serves as a legitimate culture that can be mastered to varying extents. Students who have been inculcated these cultural forms from childhood will have the greatest probability of academic success. The subsequent research aimed at testing the validity of cultural capital theory is no unified research tradition. In this article, we first discuss different understandings of the concept of cultural capital and argue that a main line of division runs between a narrow understanding, linking cultural capital to high culture, and a broader notion emphasizing the importance of more general linguistic and cognitive skills, habits, and knowledge. We argue that the formal or symbolic aspects of students’ performance, in contrast to their technical skills, are given different weight in different settings, and discuss the implications based on Bourdieu’s writings. These implications are tested out on a data set consisting of five full cohorts of Norwegian compulsory school leavers, among which three cohorts are followed through secondary school. Their school performance is recorded in great detail. We use a class scheme developed for register data aimed at capturing the basic class divisions described by Bourdieu. Many of our findings support Bourdieu’s ideas about class and culture. Among other things we find that the level of school performance varies among class factions as well as among classes on different hierarchical levels. Class inequality increases during the educational career and varies between written and oral exams.