To assess the reliability, construct validity, and predictive (concurrent) validity of the Tilburg Frailty Indicator (TFI), a self-report questionnaire for measuring frailty in older persons. Cross-sectional. Community-based. Two representative samples of community-dwelling persons aged 75 years and older (n = 245; n = 234). The TFI was validated using the LASA Physical Activity Questionnaire, BMI, Timed Up & Go test, Four test balance scale, Grip strength test, Shortened Fatigue Questionnaire, Mini-Mental State Examination, Center for Epidemiologic Studies Depression Scale, Anxiety subscale of the Hospital Anxiety and Depression Scale, Mastery Scale, Loneliness Scale, and the Social Support List. Adverse outcomes were measured using the Groningen Activity Restriction Scale and questions regarding health care use. Quality of life was measured using the WHOQOL-BREF. The test-retest reliability of the TFI was good: 0.79 for frailty, and from 0.67 to 0.78 for its domains for a 1-year time interval. The 15 single components, and the frailty domains (physical, psychological, social) of the TFI correlated as expected with validated measures, demonstrating both convergent and divergent construct validity of the TFI. The predictive validity of the TFI and its physical domain was good for quality of life and the adverse outcomes disability and receiving personal care, nursing, and informal care. This study demonstrates that the psychometric properties of the TFI are good, when performed in 2 samples of community-dwelling older people. The results regarding the TFI's validity provide strong evidence for an integral definition of frailty consisting of physical, psychological, and social domains. Copyright 2010 American Medical Directors Association. Published by Elsevier Inc. All rights reserved.