Positron emission tomography-computed tomography with F-18-fluorodeoxyglucose is widely used for evaluation of therapy response in patients with solid tumors but has not been as readily adopted in clinical trials because of the variability of acquisition and processing protocols and the absence of universal response criteria. Criteria proposed for clinical trials are difficult to apply in clinical practice, and gestalt impression is probably accurate in individual patients, especially with respect to the presence of progressive disease and complete response. Semiquantitative methods of determining tissue glucose metabolism, such as standard uptake value, can be a useful descriptor for levels of tissue glucose metabolism and changes in response to therapy if technical quality control measures are carefully maintained. The terms partial response, complete response, and progressive disease are best used in clinical trials in which the terms have specific meanings and precise definitions. In clinical practice, it may be better to use descriptive terminology agreed upon by imaging physicians and clinicians in their own practice.