BACKGROUND: In the elderly, little attention has been paid to anxiety both on a symptom dimension and as a disorder, as an independent risk factor for the incidence of activity limitations. METHODS: In a community-dwelling cohort of 1581 persons aged 65+, the association between trait anxiety symptoms (Spielberger Trait, third highest tertile) and baseline DSM-IV anxiety disorder, and 7-year incident activity limitations was determined using mixed logistic regression models. Repeated measures of activity limitations included, by increased severity level: social restriction (neighbourhood and house confined), mobility (Rosow and Breslau scale) and limitations in instrumental activities of daily living (IADL). RESULTS: Of the sample, 14.2% had an anxiety disorder at baseline. Adjusting for baseline socio-demographic and health variables, depression (past and current), antidepressant and anxiolytic drugs, baseline anxiety disorder was associated with an increased risk of incident IADL limitation (OR (95% CI): 1.84 (1.01-3.39), p=0.048) and trait anxiety with increased incidence of social restriction (OR (95% CI): 2.41 (1.42-4.09), p=0.001). Associations remained significant in participants free of depressive symptoms at baseline (OR (95% CI): 2.92 (1.41-6.05), p=0.004; OR (95% CI): 3.21 (1.31-7.89), p=0.011, respectively). LIMITATIONS: Activity limitations were self-reported and may have been over-reported in participants with anxiety. CONCLUSION: Both trait anxiety symptomatology and anxiety disorder were independently associated with increased incidence of activity limitations with a gradient of severity: trait anxiety associated with incident social restriction and anxiety disorder with more severe IADL limitations, suggesting that anxiety is a predictor of activity limitations in the elderly independently of depression comorbidity.