Establishing new surrogate end points for monitoring response to treatment is needed for current therapy modalities and for new therapeutic strategies including molecular targeted cancer therapies. PET as a functional imaging technology provides rapid, reproducible, noninvasive in vivo assessment and quantification of several biologic processes targeted by these therapies. PET is useful in a variety of clinical relevant applications, including distinguishing between radiation necrosis and tumor recurrence, determining the resectability of recurrent tumor, and evaluating response to therapy. FDG-PET has demonstrated efficacy for monitoring therapeutic response in a wide range of cancers, including breast, esophageal, lung, head and neck, and lymphoma. FDG-PET can assess tumor glucose use with high reproducibility. Following therapy, the decrease of glucose use correlates with the reduction of viable tumor cells. FDG-PET allows the prediction of therapy response early in the course of therapy and determining the viability of residual masses after completion of treatment. The molecular basis for the success of FDG-PET is the rapid reduction of tumor glucose metabolism in effective therapies. Of even higher clinical relevance is the accurate identification of nonresponders in patients without a significant change in tumor glucose metabolism after initiation of therapy. PET imaging can easily visualize these changes in metabolic activity and indicate, sometimes within hours of the first treatment, whether or not a patient will respond to a particular therapy. In contrast to CT, MR imaging, or ultrasound, PET imaging allows identification of responding and nonresponding tumors early in the course of therapy. With this information, physicians can rapidly modify ineffective therapies for individual patients and thereby potentially improve patient outcomes and reduce cost. One of the major limitations for the routine application of FDG-PET imaging for therapy monitoring is that no generally accepted cutoff values have been established to differentiate optimally between responders and nonresponders. The patient series are still relatively small and frequently consist of different tumor types and different therapy regimens. Prospective studies including a sufficient number of patients are needed to define cutoff values to differentiate between responder and nonresponder for different tumors and different treatment regimes. In the future, PET imaging can also serve in the evaluation of new therapeutic agents, new experimental treatments, and specifically in monitoring clinical phase II studies.