Suitable analogs of d-mannoheptulose are currently considered as possible tools for the non-invasive imaging of pancreatic islet insulin-producing cells. Here, we examined whether (19)F-heptuloses could be used for non-invasive imaging of GLUT2-expressing cells. After 20 min incubation, the uptake of (19)F-heptuloses (25 mM) by rat hepatocytes, as assessed by (19)F NMR spectroscopy, ranged from 0.50 (1-deoxy-1-fluoro-d-mannoheptulose and 3-deoxy-3-fluoro-d-mannoheptulose) to 0.25 (1,3-dideoxy-1,3-difluoro-d-mannoheptulose) and 0.13 (1-deoxy-1-fluoro-d-glucoheptulose, 3-deoxy-3-fluoro-d-glucoheptulose and 1,3-dideoxy-1,3-difluoro-d-glucoheptulose) µmol per 3×10(6)cells. (19)F MRI experiments also allowed the detection of 1-deoxy-1-fluoro-d-mannoheptulose in rat hepatocytes. All three (19)F-mannoheptuloses cited above, as well as 7-deoxy-7-fluoro-d-mannoheptulose and 1-deoxy-1-fluoro-d-glucoheptulose inhibited insulin release evoked in rat isolated pancreatic islets by 10mM d-glucose to the same extent as that observed with an equivalent concentration (10mM) of d-mannoheptulose, while 3-deoxy-3-fluoro-d-glucoheptulose and 1,3-dideoxy-1,3-difluoro-d-glucoheptulose (also 10mM) were less potent than d-mannoheptulose in inhibiting insulin release. The 1-deoxy-1-fluoro-d-mannoheptulose and 3-deoxy-3-fluoro-d-mannoheptulose only marginally affected INS-1 cell viability. These findings are compatible with the view that selected (19)F-heptuloses may represent suitable tools for the non-invasive imaging of hepatocytes and insulin-producing cells by (19)F MRI.