Five elementary school groups, or cohorts, were given the Test Anxiety Scale for Children and the Lie Scale for Children on three occasions over a year's time. This paper examines the results cross-sectionally and reveals between-group, or cohort effects. The youngest and oldest cohorts displayed low anxiety. Within-group longitudinal changes did not closely match this cross-sectional difference. Current data was compared to previously published longitudinal studies. Longitudinal anxiety patterns varied across cohorts and samples while defensiveness showed longitudinal decreases across all studies. The consistency in defensiveness results and the lack of consistency in test anxiety results across the three studies suggest that anxiety is more a function of situational factors and that defensiveness is more a function of organismic factors. (Author)