Cerebral function is associated with high metabolic activity that is supported by continuous supply of oxygen and glucose from the blood. Coupling between brain energy metabolism and neuronal activity has been studied extensively during the past decades. Notably, glucose metabolism was demonstrated to result from cellular cooperation between neurons and astrocytes. Although both cell types regulate synaptic transmission via the glutamate-glutamine cycle, the actual contribution of glial and neuronal oxidative metabolism is still unclear. Several research groups have tried to quantify cerebral energy metabolism in the living brain using methods, such as positron emission tomography, autoradiography and proton magnetic resonance spectroscopy (1H MRS). Yet, these methods lack chemical specificity and do not provide quantitative information specific to either cell types. The development of tracers detectable by MRS, such as 13C isotope, along with high magnetic field and dedicated hardware systems has improved sensitivity and has motivated the design of advanced metabolic modeling, in which cell types can be compartmentalized and which associated metabolic rates can be assessed independently of any prior estimations. The work of this thesis aimed at focally increasing neuronal activity and estimating the resulting changes in glial and neuronal oxidative metabolism using direct 13C MRS along with [1,6-13C]glucose infusion in vivo at 14.1 T. 4h-somatosensory stimulation paradigm was first developed in rats under alpha-chloralose. The protocol consisted on intermittently electrically stimulating both fore- and hindpaws, which resulted in a sustained and localized blood oxygenation level-dependent signal as assessed by functional magnetic resonance imaging. The relatively large activated area in the cortex allowed monitoring local 13C isotope incorporation over 4h into specific positions of glutamate, glutamine and aspartate. The analysis of the turnover curves with a two-compartment model of brain energy metabolism revealed an increase of both glial (VTCAg,+68 nmol/g/min,+22%) and neuronal (VTCAn,+62 nmol/g/min,+12%) oxidative metabolism associated with 95% increase (VNT,+67 nmol/g/min) in glutamate-glutamine cycle rate, which represents glutamatergic neurotransmission rate. The total increase in glucose oxidative metabolism was of 15% (CMRglc(ox), +67 nmol/g/min). In randomly delivering lines in 4 orientations and 2 directions, 4h-continuous stimulation of tree shrew primary visual cortex under light isoflurane anesthesia resulted in a decrease in both brain glucose concentration (-17%;-0.34 µmol/g) and phosphocreatine/creatine ratio (-9%;-0.07) after 16 minutes of stimulation onset, which recovered 21 minutes after stimulation offset. At the individual level, a nearly one-to-one relationships between VNT and CMRglc(ox) was observed. VTCAn and VTCAg were also coupled to VNT. At the group level, 20% increase in VNT (+0.038±0.042 µmol/g/min) resulted in 24% (+0.063±0.057 µmol/g/min) and 12% (+0.061±0.032 µmol/g/min) increase in VTCAg and VTCAn, respectively, resulting in 14% increase in CMRglc(ox) (+0.058±0.032 µmol/g/min). To conclude, this work demonstrates that a significant fraction of glucose is also oxidized in astrocytes and that both neuronal and astrocytic metabolism can be stimulated by and coupled to the glutamate-glutamine cycle rate.