Most sociolinguistic studies rely on apparent time, cross-sectional methods to analyze language change. On the basis of apparent time data, sociolinguists have hypothesized that cultural processes of lifespan change create predictable cycles of linguistic behavior in which adolescents lead in the use of vernacular variants and advance sound change (Eckert 1997). While adolescence is hypothesized to be central to vernacular optimization and language change processes, only longitudinal studies reveal whether individuals change their linguistic behavior in predictable ways across adolescence. Furthermore, longitudinal data about individual trajectories of change allow linguists to confirm or disconfirm apparent time data. As a longitudinal study of over 67 African Americans from infancy to post-high school, the Frank Porter Graham (FPG) study presents a unique opportunity to document language variation across the lifespan. This analysis is the first longitudinal acoustic analysis of vocalic variation from childhood to early adulthood. Because African American English (AAE) vowels in the Piedmont region of NC are stable, this study can explore the extent to which life-stage variation influences participation in ethnolinguistic vowel systems without the confound of a change in progress. Additionally, because longitudinal trajectories of AAE morphosyntactic/consonantal variables are documented, comparisons across linguistic subsystems reveal the extents and limits to which life-stage patterns predict linguistic cycles of behavior. This study focuses on a subset of 20 individuals at approximately ages 9, 12, 15, and 20. Although all participants are from the Piedmont region of NC, individuals come from two communities with different demographics. Hierarchical regressions show that, while participation in AAE vowels strongly correlate with community and school demographics, stable vocalic variables do not undergo aggregate-level peaking patterns consistent with age-grading. Instead, stable aggregate patterns camouflage idiosyncratic individual trajectories. A lack of group patterns for vowel variation across adolescence suggests that life-stage variation does not affect all linguistic systems equally; age-grading is a minority pattern perhaps associated with stereotyped features and/or morphosyntactic/consonantal variables. Because age-grading is not a predominant pattern for non-stereotyped vocalic variation, apparent time peaks in adolescent vowel data should not be taken for granted as a default product of age-grading.