The lumped constant (LC) that relates the steady-state phosphorylation rate of 2-[18F]-fluoro-2-deoxy-D-glucose (2-FDG) to that of glucose was determined in an isolated working rat heart model by direct assay of phosphorylation product formation. Five conditions were tested: 5 and 30 mM glucose without insulin, and 2, 3.5, and 5 mM glucose + 10 mU/ml insulin, all at high external work load. Hearts were continuously perfused with 2-FDG and tritiated glucose without recirculation. The steady-state production of tritiated water was used to monitor the glucose phosphorylation rate. Perfused hearts were freeze-clamped and extracted in perchloric acid, and 2-FDG-6-phosphate was separated from 2-FDG with a formate column. The accumulation of 2-FDG phosphorylation products in tissue was also determined from the slopes of the total tissue radioactivity time courses measured by external gamma-ray detection. Without insulin, the LC value decreased 18% as perfusate glucose concentration was increased sixfold (0.94 +/- 0.06 at 5 mM vs. 0.77 +/- 0.17 at 30 mM). With insulin, the LC rose from 0.33 +/- 0.03 at 5 mM to 1.19 +/- 0.05 at 2 mM glucose concentration. The trends can be interpreted in terms of the concept of control strength; the LC value rises as glycolysis becomes rate limited by transport into cells. This potential variability of the LC must be addressed in the quantitative interpretation of myocardial deoxyglucose studies.