Preserved glucose metabolism in ischemically injured, dysfunctional myocardial tissue as demonstrated on PET imaging predicts functional improvement after revascularization. To characterize more precisely the relationship between regional myocardial blood flow, viability and extent and severity of flow and metabolism abnormalities, we developed a PC-based semiquantitative analysis technique using 13N-ammonia and 18F-deoxyglucose polar map displays. A data base for mean values (m) and standard deviations (s.d.) for relative 13N activities reflecting regional myocardial blood flow, relative 18F activities normalized to normal flow regions reflecting regional glucose utilization and the difference of normalized 18F and 13N activities as an index of a flow-metabolism mismatch was established in 11 normals. Parametric polar maps were derived by comparing patient data to a normal range defined as greater than m - 2 s.d. for relative myocardial blood flow and less than m + 2 s.d. for both relative glucose utilization and the difference between normalized 18F and 13N activities. Semiquantitative indices of extent and severity of blood flow defects, of relative increases in glucose utilization and of flow-metabolism mismatch areas are generated for the entire myocardium and the three coronary territories. The approach promises to be clinically useful to confirm presence and absence of flow and metabolic abnormalities and to assess their extent as a potential predictor of functional outcome after therapy.