Life expectancy increasing and cancer incidence rising with age, geriatric and cancer care will become a significant medical, public health, challenge. It is possible that the lack of efficacy of cancer therapies in the elderly may simply be due to the fact that physicians reduce anticancer drug doses empirically, in order to avoid "putative" toxicities that might arise as a result of alterations of physiological functions or as a result of co-morbidities generally present within this population. However, many authors have demonstrated that some patients over 70 years old could tolerate and obtain same benefit from therapies as younger adults, when some who are frail need less aggressive therapies. It is imperative that decisions regarding the management of cancer in older persons should be based upon the individual needs and fitness of the patient, and not based on chronological age. The main difficulty is the lack of scientific references on optimal treatment-dosing in the elderly, and we cannot extrapolate, as it stands, the information from younger patient. Indeed, aging is associated with multidimensional changes, including alteration of physiological status, comorbidities and polymedications. These changes may lead to pharmacokinetic (PK) modifications, namely in absorption, distribution, metabolism and elimination of drugs as well as pharmacodynamic (PD) modifications. Prospective studies need to be implemented in the elderly in order to study in depth the PK and PD properties of the drugs used for these patients, and establish PK-PD relationships in this specific population. Such studies have been successfully conducted in the elderly, some of them leading to dose recommendations. This paper detail the different sources of PK-PD variability in the elderly, some practical considerations regarding the design of studies using the population approach, as well as some examples of studies performed in the elderly. We conclude with some recommendations in this population at risk.