Although it is well-documented that hyperoxia at 100% inspired O₂ leads to decreased cerebral blood flow in humans¹, reports on where vasoconstriction occurs along the cerebrovascular tree are limited^2,3. Due to the potential of hyperoxia as treatment for cerebral ischaemia⁴, it is important to fully understand its effects on the human cerebrovasculature. Therefore, we aim to investigate the effect of hyperoxia along the cerebrovascular tree with short inversion time (TI) arterial spin labelling (ASL)5 healthy volunteers (2 F, 27.2 ± 1.6 years) underwent a MRI scan session (3 T, GE HDx, Milwaukee, USA). Single slice PICORE ASL images were acquired: single inversion time (TI = 450ms), TR = 600ms, TE = 3 ms, number of time points = 3200, slice along the anterior-posterior cingulate line, resolution = 3x3x7 mm³, FOV = 198 mm, label thickness = 50 mm (10 mm below the imaging plane). Images were acquired for 8 different levels of fixed inspired O₂ (FiO₂) ranging from 21% (normoxia) to 42% in 3% increments. Each level lasted 4 minutes and was created by mixing medical air and 100% O₂. Automated flow controllers driven by in-house Matlab (R2012b) scripts ensured a randomized order. Gases were delivered via a tight fitting facemask. To determine baseline arterial blood volume (aBV), a multi-TI PICORE ASL scan was acquired at normoxia. Parameters: imaging and labelling planes identical to the above scan; 8 TIs (150ms to 850ms, spacing 100ms); variable TR (minimized); single TE = 3ms; 320 time points (20 tag-control pairs per TI). Throughout the scans, physiological monitoring measured end-tidal partial pressures of O₂ and CO₂ (P_ETO₂ and P_ETCO₂), and cardiac and respiratory cycles. A running subtraction (ΔM) of tag and control images was performed for the short TI time series. Note that, due to the short TI, we assumed that all arterial blood is labelled and therefore ΔM is proportional to the aBV in the voxel. To obtain average ΔM per FiO₂ the parameter maps were averaged over the last minute of each hyperoxia level, i.e. when P_ETO₂ was stable. Maps of aBV were created by voxelwise fitting of an arterial input function⁵ to the average tag-control subtraction images from the multi TI baseline scan at normoxia. Masks of large (aBV_Large = aBV > 5 %_v), medium (aBV_Medium: 2 %_v < aBV < 5 %_v), and small arteries (aBV_Small: 0.5 %_v < aBV < 2%_v) were created (see Fig 1). Repeated measures ANOVA (RM-ANOVA) was used to investigate the effect of Hyperoxia and Artery Size on ΔM.The RM-ANOVA showed a significant interaction between Hyperoxia and Artery Size, across all voxels (F(14,56) = 8.3, p < .001). Separate RM-ANOVA for different artery sizes showed that there was a significant decrease in ΔM with oxygenation level for the large arteries (F(7,28) = 7.8 , p < .001), while there was no effect of oxygenation on ΔM for aBV_Medium (F(7,28) = 2.4, p = 0.07), and a significant increase in ΔM for aBV_Small­­ (F(7,28) = 6.1, p < .001). Fig 2 illustrates these results. Note that the decrease in ΔM in aBV_Large at FiO₂ = 42% (16.4 ± 8.4 %) is greater than the 2% decrease in ΔM expected based on shortening of T₁ with increased oxygenation¹. There was no significant effect of oxygenation level on P_ETCO₂ F(7,35) = 0.9, p = 0.47), which rules out hypocapnia affecting vascular tone. Our results suggest that the effect of hyperoxia on cerebral vascular tone is dependent on the size of the artery. The induced vasoconstriction in large cerebral arteries is in line with the role of oxygen in the signalling pathway between glial and smooth muscle cells². The absent response in aBV_­Medium and apparent vasodilation in aBV_Small­ are not easily explained by current literature, although this work does match previous results of not finding a change in aBV with isocapnic hyperoxia in whole brain grey matter, effectively summing the response seen here across all arteries³. A possible explanation is an autoregulatory mechanism to maintain cerebral blood flow; in response to the increased resistance of the larger arteries, the more distal small arteries dilate to maintain perfusion (assuming that at moderate hyperoxia changes in perfusion are negligible¹). Further investigation of the differential response of the cerebrovasculature to hyperoxia is required to fully understand the effects of oxygen on cerebral vascular tone.