Perfusion-weighted imaging studies have demonstrated that there is a widespread cerebral hypoperfusion in patients with MS, regardless of the clinical subtype [1]. The mechanism and the role of hypoxia are still unclear. Recently, several studies investigating the presence of hypoxia in Experimental-Autoimmune-Encephalomyelitis (EAE), an animal model of MS which affect primarily the spinal cord (SC), have been reported. These studies used susceptibility weighted MRI [2] or invasive techniques like insertion of an oxygen-sensitive probe [2]. However, to the best of our knowledge no quantitative measurements of the blood flow have been reported so far. Arterial Spin Labelling (ASL) is a valuable tool in the investigation and understanding of CNS pathologies, where perfusion plays a key role. While it has been applied extensively to the brain, very few SC ASL studies have been reported due to the inherent experimental challenges [4,5]. The purpose of this work was to longitudinally investigate the SC blood flow (SCBF) during the different phases of disease progression in EAE rats using an optimized ASL technique.Experiments were performed on Dark Agouti rats. EAE was induced by injecting a recombinant myelin oligodendrocyte protein (rMOG) in incomplete Freund’s adjuvant (IFA). Control animals received an emulsion of IFA and saline. They were weighed and assessed daily for functional deficit on a 10-point scale [2]. The animals were scanned before immunization and at different stages of the disease: first peak (12-17 days post-immunization (dpi)), remission (16-19dpi), and relapse (19-23dpi) (Fig.1). MR data were acquired on a 9.4T Agilent scanner using transmit volume coil and two element receive array coil placed at the lumbar level. Six animals (3 controls and 3 EAE) were scanned under anaesthesia. Images with an in-plane spatial resolution of 125x125um2 were obtained using an optimised ASL sequence with adiabatic RF pulses and reduced FOV [4]. Four 2mm slices, with 1mm gap, were acquired using four-shot-EPI. ASL labelling was performed using a pre-saturation-FAIR-Q2TIPS sequence [6] with a recovery time of 3.2s and TI1/TI2=1.55/1.65s, TE/TR=20/5015ms and an acquisition time of 25min.Typical SCBF maps at the lumbar level are depicted in Fig.3, in which highly perfused GM (relative to WM) can be noticed. The average SCBF of the healthy animals (before immunisation, n=6) in GM and WM was 382.0±14.8ml/100g/min and 101.4±16.3ml/100g/min, respectively. Each of the control animals (n=3) was scanned at three time points. The averages of the nine measurements were SCBFGM=392.1±19.2ml/100g/min and SCBFWM=80.8±17.9ml/100g/min. At the first peak of the disease (average scores of 8), SCBFGM (234.4±35.3ml/100g/min) decreased by 38.6% compared to the baseline (p=0.007). Compared to these values, SCBFGM (299.4±21.9ml/100g/min) increased by 27.7% during the brief remission period (average scores of 3.5, n=2) but 21.6% lower than baseline measurements (p=0.04). One of the EAE rats was not scanned during that period because it relapsed earlier than expected. Subsequently, SCBFGM (216.9±43.0ml/100g/min) decreased drastically by 27.6% (43.2% compared to the baseline, p=0.008) during the relapsing phase (average scores of 9). One of the EAE rats which showed slight functional deficit improvements (score=7) was scanned a second time during the relapsing period. SCBFGM (315.2±11.5ml/100g/min) increased by 45.4% compared to the previous time point (17% lower than baseline). Except the remission time point, no significant differences were noticed between controls and EAE within the WM.The quality and spatial resolution of the SCBF maps permitted an accurate identification of the highly perfused GM and less well perfused WM. The baseline measurements showed high group reproducibility (3.9% and 16.3% with GM and WM, respectively). The relatively higher variability with WM is due to the lower SNR compared to GM. SCBF of the different control animals and at different time points was highly reproducible. The variabilities were 4.9% and 22.1% within GM and WM, respectively. Measured SCBFGM of EAE rats was significantly lower than controls. After an initial drop, SCBF significantly recovered during remission before drastically decreasing. These measurements demonstrated for the first time on EAE animal model that the neurological deficits are strongly correlated with impaired blood flow. These preliminary findings, which need to be validated on a larger animal cohort, are in agreement with previous clinical studies. This decrease could be due to massive infiltration and extravasation of inflammatory cells associated with pronounced edema which together could obstruct outflow or mechanically compress vessels of a swollen cord. Reduced SCBF might be mediated by enhanced blood levels of the potent vasospastic peptide endothelin-1 as it was observed on patients with MS [1]. Further studies are needed to investigate the causes and consequences of hypoperfusion/hypoxia in MS.