Over the past decade, clinical pharmacokinetic studies in the elderly, both healthy subjects and patients, have burgeoned. This data base has provided a useful first approximation to the understanding of age-related changes in drug disposition. It is now appropriate to reflect on the approach for future studies. The most accessible population, healthy elderly who are otherwise drug-free, may not provide completely relevant data for extrapolation to the very elderly, concurrently medicated patient population. Similarly, the most straightforward study, single dose drug exposure, may provide an incomplete understanding of pharmacokinetics during multiple dose drug administration in the very elderly patient. With careful attention to details of the study and informed consent procedures, it is appropriate to obtain the needed data.