Objective: To examine long-term outcome of late-life anxiety disorders and utilization of mental health care services. Method: A cohort of subjects (aged ≥55 years) with an anxiety disorder (n = 112) was identified in the Longitudinal Aging Study Amsterdam (n = 3107). At 6 year follow-up, the rate of persistence and prognostic factors for persistence of anxiety were established. Results: Six years after baseline 23% of our sample met the criteria for an anxiety disorder. Another 47% suffered from subclinical anxiety symptoms. Persistence of anxiety was associated with a high score on neuroticism at baseline. Use of benzodiazepines was high (43%), while use of mental health care facilities (14%) and anti-depressants (7%) remained low in those with persistent anxiety. Conclusion: Results indicate that those high in neuroticism are at greater risk for persistence of anxiety. Efforts to enhance appropriate referral of anxious older adults do not seem to have had the desired effect. © Blackwell Munksgaard 2005.