In an analysis of the consistency of self-reported age at first intercourse using longitudinal data from the National Youth Survey, 28-32% of adolescents reported an age at first intercourse inconsistent with the information they provided up to seven years later as adults. Overall, while females were the most likely to offer consistent responses (70%), while black males were the least likely to do so (27%). Multivariate analyses indicated that in addition to race and gender, some social and economic factors were significantly associated with inconsistent reporting. For example, those who lived in a two-parent household were less likely than those from a one-parent family to report an earlier age at first intercourse as adolescents than they reported as adults. After controlling for these inconsistencies, overall predictors of adolescent sexual behavior remained unchanged. Although the analysis could not determine which time period reflected more accurate data, it does suggest limitations to using self-reported data to construct estimates of adolescent sexual activity and to evaluate programs designed to alter adolescent sexual behavior.