Aims: The frailty concept is of growing improtance. There is a need for the development of a generic instrument to fasten dwon clincians' judgment about the autonomy and the degree of frailty of the elderly. It can be a worthwile tool in the early detection of dementia in primary heatlh care. Methods: The validity of the Frial instrument was tested. Out of the Qualidem study population, for 553 persons a Camdex-R score and MMSE are available to test the diagnostic parameters. Second, using data of 845 subjects at home and in the homes for the elderly, validity has been studied by the non linear explorative principal component analysis using Princals. Results: The diagnostic parameters of the Frail test in single use and in combination with MMSE for 553 persons are good. Sensitivity is good but specificity is weak. A good Oddsratio is significantly higher when it is used combining Frail with MMSE. The three dimensions of the PRINCALS analysis explained 72% of all variance. The component loadings express the strenght of association between each of the items on the one hand and the identified cominent (i.e. dimension) on the other. Conclusion: The Frail insturment has a good diagnostic value. There are strong arguments for a good content validity of the instrument.