Background: There is no consensus on how to define subthreshold anxiety. Based on functional impairment, we aim to evaluate the use of a diagnostic instrument and an anxiety severity questionnaire to derive an empirical cut-off for defining clinically relevant, subthreshold anxiety. Methods: Our sample consisted of 1788 subjects without full-syndromal anxiety disorders from the Netherlands Study of Depression and Anxiety (NESDA). We used ANCOVA to compare the Composite International Diagnostic Interview (CIDI) and the Beck Anxiety Inventory (BAI) in their association with functional impairment, measured by the World Health Organization Disability Assessment Schedule (WHODAS II). The BAI was selected over the Fear Questionnaire (FQ) and the Penn State Worry Questionnaire (PSWQ) for its highest associations with anxiety disorders. ROC analysis determined the percentage of functionally impaired with subthreshold anxiety based on found cut-offs. Results: The CIDI was very modestly associated with functional impairment, possibly because of skip rules, wording or scoring, while the BAI was highly correlated to functional impairment. A score of 11 on the BAI defined clinically relevant subthreshold anxiety. This cut-off identified 36.0% of the most functionally impaired as having subthreshold anxiety. Limitations: No "natural" cut-off on the BAI for determining subthreshold anxiety could be determined. The cut-off point of 11, based on a large effect size relative to normal anxiety, thus remains somewhat arbitrary. Conclusions: The questionnaire BAI is more suitable than the interview CIDI to measure clinically relevant, subthreshold anxiety. The BAI score of 11 appeared to be the most appropriate cut-off for identifying clinically relevant subthreshold anxiety. © 2010 Elsevier B.V. All rights reserved.