Frailty is a commonly used term indicating older persons at increased risk for adverse outcomes such as onset of disability, morbidity, institutionalisation or mortality or who experience a failure to integrate adequate responses in the face of stress. Although most physicians caring for older people recognize the importance of frailty, there is still a lack of both consensus definition and consensual clinical assessment tools. The aim of the present manuscript was to perform a comprehensive review of the definitions and assessment tools on frailty in clinical practice and research, combining evidence derived from a systematic review of literature along with an expert opinion of a European, Canadian and American Geriatric Advisory Panel (GAP). There was no consensus on a definition of frailty but there was agreement to consider frailty as a pre-disability stage. Being disability a consequence rather than the cause of frailty, frail older people do not necessary need to be disabled. The GAP considered that disability (as a consequence of frailty) should not be included in frailty definitions and assessment tools. Although no consensual assessment tool could be proposed, gait speed could represent the most suitable instrument to be implemented both in research and clinical evaluation of older people, as assessment of gait speed at usual pace is a quick, inexpensive and highly reliable measure of frailty.