Discussion surrounds the question as to whether criteria for generalized anxiety disorder (GAD) should change, particularly in youth. This study examines the effects of possible criteria changes on GAD prevalence and clinical correlates. DSM-IV GAD was assessed using the M-CIDI in a community sample of adolescents and young adults. Diagnostic thresholds were modified in two age spans (9-20 and 21-34 years) using a person-by-year data file (N = 38,534 cases). Relaxing the duration or excessiveness criteria led to the most pronounced changes in GAD prevalence, while relaxing frequency, uncontrollability, or associated-symptom criteria had smaller effects. A lower duration requirement increased rates more in older than younger age spans. Opposite effects occurred for changes in associated-symptoms or clinical-significance criteria. Broader GAD definitions identified cases in both age spans that appeared mostly milder than DSM-IV cases but that still differed from non-GAD cases in various clinical factors and validators. Developmental aspects require stronger consideration in future diagnostic systems.