The physiological and biochemical measurements that can be performed noninvasively in humans with modern imaging techniques offer great promise for defining the precise state of a patient's disease and its response to therapy. In general, there are two critical points in drug development when PET measurements are likely to be particularly useful: (1) In preclinical studies, a new drug can be precisely compared to standard therapies or a series of analogs can be screened for further development on the basis of performance in appropriate animal models. (2) In phase I-II human studies, classic pharmacokinetic measurements can be coupled with imaging measurements (a) to define optimal dosing schedule; (b) to define the potential utility of interventions in particular clinical situations; and (c) to formulate the design of phase III studies that are crucial for drug licensure. In general, the types of measurements that are possible can be grouped into the following categories: 1. In those situations in which the drug can be radiolabeled, the time course of tissue delivery can be determined noninvasively in vivo in health and disease. Such information should be useful for determining dosing schedules, establishing efficacy, and predicting possible toxicity. 2. Ligand-receptor binding can be assessed in vivo in two ways. The ability of the drug to displace standard radiolabeled ligands from their receptors can be determined; alternatively, labeled drug can be used to more directly assess the distribution and time course of binding. These measurements are particularly useful for studying drugs that are active in the central nervous and cardiovascular systems. 3. Measurements of tissue metabolism will be useful in determining the effects of therapies aimed at particular metabolic abnormalities. In addition, these measurements may be useful in defining viability and function of tissues in such widely disparate clinical situations as cancer chemotherapy and cardiology. For example, effects of CNS or cardiovascular drugs can be monitored by observing 18FDG metabolism in brain and heart. We suggest that the joining of classic clinical pharmacology to exquisite imaging measurements will help form the basis for 21st-century clinical drug development.