FID based spectroscopic imaging (SI) provides full signal intensity and minimizes chemical shift displacement errors for data acquired at high field [1]. Using this method, high spatial and temporal resolution can be achieved (1ul voxel size, 12 min) enabling functional metabolic measurements in mice using various stimulation paradigms such as pharmacological or sensory stimulation [2, 3]. However, for such measurements the shim gradients must be optimized on small volumes typically not exceeding 40 µl on the mouse brain; this limits SI data acquisition to typically one slice only. In order to increase the volume coverage i.e. allow for the simultaneous measurement of several slices, we implemented a dynamic shimming approach thereby optimizing the magnetic field homogeneity for each slice individually, warranting the spectral quality required for accurate analysis of fMRS data.All experiments were carried out using a BioSpec 94/30, (AVIII HD, PV6.01, Bruker BioSpin MRI GmbH, Ettlingen, Germany) small animal MR system operating at 400 MHz. A four-element receive-only cryogenic phased array coil (2x2 geometry, overall coil size 20x27mm2) was used in combination with a linearly polarized room temperature volume resonator for transmission. All in vivo experiments were carried out in strict adherence with the Swiss law for animal protection. Mice (N=3) were anesthetized using isoflurane (1.5%); intubated and artificially ventilated with an oxygen/air (20% / 80%) mixture. Two axial slices were selected comprising somatosensory cortex, striatal regions and ventricles in the first slice and mainly thalamic and cortical regions in second slice. Optimal first and second order shim terms were evaluated for a volume of 36 µl within each slice separately prior to the experiment using field maps. The shim sets were stored and updated for each slice in the pulse sequence together with the corresponding working frequency. For SI, following parameters were used: TR: 2500ms; FOV: 1.5x1.5cm; matrix: 17x17; slice thickness: 1.3mm; acquisition time: 12min. Scans were performed using VAPOR water suppression interleaved with six saturation slices for outer volume suppression. SI scans were repeatedly acquired, alternating baseline and stimulation periods. Stimulation paradigm: 2mA, 5 Hz, 40s stimulation, 20s rest, 10 cycles (for a total duration of 10 min). Relative quantification was performed using LCModel [4].The volumes of interest (or shim volumes) and associated spectra extracted from a cortical regions in the first slice and a thalamic region in the second slice are displayed in Fig. 1. A line width of 10 Hz on average was achieved in both spectra using dynamic shim updates (red line), while at least twice as large values (22 Hz) were observed for spectra of the same locations when using a static shim volume comprising both slices. This resulted in improved spectral resolution and an SNR increase by a factor of two on average (Fig. 2A, B). In general, similar spectra quality has been observed in both slices using dynamic shimming compared to measuring each slice individually (data not shown). Metabolites involved in general energy metabolism and neurotransmission such as Glutamate (Glu), Glutamine (Gln), GABA and lactate (Lac) could be quantified with Cramer-Rao lower bounds (CRLB) below 15% when using dynamic shimming, while significantly larger CRLB values have been obtained when using a static global shim (Fig. 3). Preliminary experiments revealed that upon electrical stimulation of the right hind paw, increased levels of Lac and Glu where measured in the left somatosensory cortex (Fig. 4A) as previously reported [3], while no prominent changes have been found in thalamus (Fig. 4B).Metabolic changes associated with stimulus evoked neural activity can be measured using functional SI if the data quality is sufficient. Due to high demands on magnetic field homogeneity, shim gradients are generally optimized in a single slice restricting the volume coverage to few regions of interest. Implementing a dynamic shim update approach allows larger brain coverage and therefore a direct comparison of several brain regions involved in the processing of e.g. sensory stimuli (e.g. thalamus and cortex). Metabolic changes upon stimulation when acquiring spectra in a multi-slice experiment have been found comparable to outcomes of previous studies using shim volumes restricted to a single slice in combination with static shim procedures.