To examine the oxygen-dependence of glucose consumption in solid tumors, we monitored gradients of glucose, lactate, and hypoxia in R3230Ac and FSA tumors growing in Fischer 344 rats. Bioluminescence imaging, detection of Hoechst 33342, and immunostaining of the hypoxia marker EF5 [2-8-N-(2,2,3,3,3-pentafluoropropyl)acetamide] were done in serial tumor slices. Glucose and lactate levels were also determined in liver and blood. Cells were further tested for glucose consumption and lactate production in vitro. In both tumor types, EF5 staining indicated similar maximum levels of hypoxia; the most intense staining occurred in perinecrotic regions. Glucose concentrations were highest in liver, declined from blood to tumor edge, further into vital tumor regions, and were lowest close to necrosis. Glucose was significantly lower in FSA than in R3230Ac tumors. Glucose concentrations in R3230Ac tumors were consistently higher in nonhypoxic than in hypoxic areas, with maximum values equal to systemic blood levels. Glucose in FSA tumors was close to zero, regardless of the presence or absence of hypoxia. Lactate did not differ significantly between the tumor types. FSA cells in culture showed a trend towards higher aerobic glucose consumption versus R3230Ac. Both cell lines increased their lactate production to similar levels under hypoxia. We conclude that both R3230Ac and FSA tumors retain the Pasteur effect, i.e., hypoxia triggers increased glycolysis. However, our results imply that increased aerobic glucose utilization leads to low glucose levels in FSA and a situation where supply limits uptake. This explains the repeatedly observed correlation between tumor blood flow and 18 Fdeoxyglucose uptake. (Cancer Res 2005; 65(12): 5163-71)