Arterial transit time (ATT) reflects the time taken for the arterial spin labeling (ASL) tracer to travel from the labeling plane to an imaging voxel. This parameter reflects the hemodynamic status of the cerebral vasculature and has added value to conventional cerebral blood flow (CBF) measurements in ASL^1-3. One approach to measure ATT with ASL^{4, 5} is Flow Encoding Arterial Spin Tagging (FEAST)^{6, 7}. Here, we investigate associations between various ASL and FEAST-based metrics in a large sample of elderly with hypertension. It has previously been shown that spatial coupling of high macro-vascular signal and low tissue signal in a single-PLD CBF image is an indication of prolonged ATT, which is problematic for CBF quantification but nonetheless carries clinical information reflecting cerebrovascular attributes². We hypothesize that macro-vascular contamination of CBF will contribute to an increased spatial Coefficient of Variation (CoV) of a single mean CBF map ‒ not to be confused with temporal signal fluctuations. E.g., if the ASL tracer has not arrived in the tissue, the CBF signal intensity difference between vascular and tissue regions will be high, whereas if the ASL tracer has arrived in the tissue then the CBF signal intensity will be spatially homogeneous (Figure 1). The aims of the current study were to 1) relate the spatial CoV with other ASL-derived metrics and 2) explore potential associations with vascular risk factors.194 community-dwelling elderly (46% male, aged 77.4 yrs (SD 2.5)) were selected from the Prevention of Dementia by Intensive VAscular care (PreDIVA) study⁸. Two consecutive GE-EPI background suppressed pseudo-continuous ASL scans, with and without vascular crushing (Venc, 5 cm/s; b, 0.6 s/mm2) were acquired ‒ labeling duration = 1650 ms, PLD = 1525-2080 ms and NSA = 20 for each scan. ATT maps were calculated using the FEAST equation, with the ratio between crushed (C CBF) and non-crushed CBF (NC CBF) as input^{6, 7}. C CBF and NC CBF were quantified with a single compartment model⁹. 3D-T1 scans were segmented and non-linearly registered to MNI space using SPM12 and DARTEL; the same transformations were applied to the CBF maps⁶. The spatial CoV was calculated as SD CBF/mean CBF spatially across all GM voxels for an individual subjects CBF map. Linear regression analyses were performed to explore associations between the various ASL metrics and vascular risk factors age, gender and body mass index (BMI).Visually, the CoV was related to the heterogeneity in the CBF and ATT maps. Note the relative homogeneity of the mean NC CBF maps for the 10 subjects with the lowest GM CoV (Figure 1a) compared to the heterogeneous and angiographic-like appearance of the mean NC CBF maps for the 10 subjects with the highest GM CoV (Figure 1b). Although this difference is still visible, it is less pronounced for C CBF. The size of the cortices and space between the cortices on Figure 1 suggest that subjects with high CoV have less GM mass than subjects with low CoV (Figure 1b vs. 1a, 3rd row). Figure 2 displays scatter plots and correlation values between the ASL-based measures, which all corresponded to p<0.0001, except for the correlation between ATT and NC CBF (p=0.3). ATT was associated with gender (men having longer ATT, t=-6.2, p<0.001), NC CoV (t=2.4, p=0.019) and C CBF (t=-3.0, p=0.004) but not with age nor with NC CBF (p>0.9). Furthermore, BMI was associated with ATT (t=-2.2, p=0.026) and NC CoV (t=3.1, p=0.002) but surprisingly it was not associated with C CBF nor NC CBF in this sample (p>0.7).Our results show that ATT increases linearly in relation to the non-crushed CBF CoV estimate, representing a “free-lunch” method to assess ATT effects on CBF images. Furthermore, our results suggest that the signal variability between macro- and micro-vascular components in the NC CBF scans are related to ATT, which are minimized by crushing. These results show the potential of the spatial CoV in studies where ATT is not acquired. An advantage of the spatial CoV in elderly could be its relative insensitivity to the signal-to-noise ratio of tissue perfusion, if macro-vascular signal is still suspected despite optimized ASL acquisition parameters. This could explain the stronger association with BMI that we observed for spatial CoV compared to ATT, as well as the normal histogram shape of CoV compared to the skewed appearance of the ATT histogram. Therefore, our data suggest that ATT ‒ as estimated by the spatial CoV ‒ might serve as a global ASL-based biomarker of cerebrovascular disease, with relevance to vascular dementia.