Underuse is defined as the absence of initiation of an effective treatment in subjects with a condition for which one or several drug classes have demonstrated their efficacy. Indeed, "effective treatment" actually means favourable benefit/risk ratio. To propose a detailed and functional definition of underuse for frail elderly we should discuss, beforehand, the better way to assess benefit/risk ratio of drugs in this population. Our work is based on a literature review in the field of inappropriate prescription and therapeutic optimization. We can foresee the hard way to accurately define underuse for frail geriatric patients because of the difficulties encountered to demonstrate drug efficacy, drug effectiveness, or even more drug risk in this specific population. Potential benefit of underused medications in this population are poorly evaluated before and even after market authorization. Premarketing clinical trials and pharmacovigilance also yield only relatively restricted information on safety of use. Underuse is a non optimal prescription modality and presumes a judgment on prescription act. This can lead to recommendations or quality indicators. It should therefore be scientifically valid and closely fit with a loss of health or loss of quality of life with a satisfying proof level. But the literature generally adopts an unsophisticated point of view. Medication introduction on the basis of a debatable definition of underuse could lead to an accumulation of useless drugs with potential adverse effects which is overuse.