We compared demographic and comorbidity profiles of subjects with varying levels of anxiety pathology to test if the clinical characteristics of generalized anxiety disorder (GAD) subjects differed from those of other subjects with generalized anxiety. Using Diagnostic Interview Schedule data from the 1993 follow-up study of the Baltimore cohort of the Epidemiologic Catchment Area Program, we divided subjects into the following five mutually exclusive symptom categories: a) DSM-III-R GAD; b) 6 months of worry or anxiety with six associated symptoms; c) 1 month of anxiety with or d) without six symptoms; and e) no anxiety. The first three groups were homogeneous with regard to demographic and comorbidity profiles, but their profiles differed from those of subjects with no anxiety or fewer than six symptoms. Thus, requiring six symptoms produced a group with a particular epidemiologic profile. Neither the nature of the subjects' worries nor the duration of symptoms influenced this profile. These results are discussed in terms of GAD's construct validity.