Longitudinal surveys are commonly understood to be more effective than retrospective ones at capturing changes in practices and representations over the individual life course because they reduce bias due to memory gaps and the “biographical illusion” (Bourdieu’s term). This article studies precisely those types of bias, discussing the contributions and limitations of the longitudinal approach on the basis of findings from a French Ministry of Culture survey entitled “Cultural repertoires of children and adolescents” (“Univers culturels des enfants et des adolescents,” 2002–2008). We analyze inconsistencies in adolescents’ consecutive responses to questions about their cultural leisure outings. Inconsistencies are often treated as errors to be corrected, but it is highly instructive to submit them to sociological analysis and examine how they fit into the process of individual life history construction.