To increase our understanding of cancer and improve cancer treatment on a rational basis we need to identify both qualitative and quantitative differences between normal and neoplastic tissue. The multimodality approach to cancer treatment includes radiotherapy, chemotherapy, hyperthermia, and immunotherapy. Most of the data on which we base our therapeutic strategies have been derived from in vitro studies or animal tumor models. More information is required on the physiology of in vivo human tumors and their response to therapy. Positron emission tomography allows the regional tissue concentration of a positron emitting radionuclide to be measured in absolute units. If valid tracer models can be formulated that accurately describe the fate of an administered "biological" tracer then the physiological process under investigation can be measured quantitatively. The sequential inhalation of C15O2, 15O2, and 11CO allows regional tissue blood flow, oxygen utilization and blood volume to be measured in absolute units. Tissue perfusion, a measure of nutrient (eg, oxygen) supply, drug delivery, or a means of heat dissipation, is of immense importance to oncologists. The oxygen-15 technique has been used not only to study regional blood flow and oxygen utilization in both tumor and normal tissue but also their response to therapeutic intervention. In those studies were tracer models are thought to be less than complete (eg, due to insufficient biological data) then only a semiquantitative or qualitative assessment of the pathophysiological state may be possible. In this respect, tumor function has been characterized by the rate of uptake of 18F-2-deoxyglucose. This technique has provided a means of tumor grading and differentiating between radiation-induced tissue necrosis and tumor recurrence. Metabolic imaging with labeled amino acids appears particularly useful in the delineation of tumor extent. Blood brain barrier integrity and the pharmacokinetics of cytotoxic drugs have also been studied quantitatively. In the future, receptor studies are likely to play an increasing role. In this review dealing with the positron emission tomography oncologic work to date, emphasis has been placed on the physiological information obtainable and how it may further our understanding of cancer and its treatment.