Increasing the concentration of oxygen dissolved water is known to increase the recovery rate (R1) of longitudinal magnetization (T1)[1-3]. Recent work exploiting changes in the T1-dependent steady state MR signal have relied on indirect O₂-dependent T1 changes, induced via high flow hyperoxic gas mixtures, as potential surrogate markers of tissue oxygenation status [4]. Nevertheless, direct T1 changes in response to precise hyperoxic gas challenges have not yet been quantified and the actual effect of increasing arterial oxygen concentration ([O₂]_A) on tissue T1 remains unclear. The aim of this work was to use native T1 (qT1) mapping to measure tissue T1 changes in response to precisely targeted hyperoxic respiratory challenges.5 Healthy subjects (2 female, 35.5±10.5 years) were scanned on a Philips 7 tesla MRI scanner using a 32 channel TX coil. End-tidal O₂ pressure (PetO₂) was modulated (under normocapnia) using a RespirAct (Thornhill Research, Toronto). Separate whole brain qT1 maps (sequence information in [5]) were acquired throughout four targeted PetO₂ levels: baseline ~ 110 mmHg, 250 mmHg, 350 mmHg and 500 mmHg). Scan parameters were as follows: multi-slice FFE with EPI readout, TR/TE: 10000/9.42 ms, scan duration: 4 min, voxel dim: 1 x 1 x 1.5 mm, slices: 46, FOV: 224 x 224 x 91.5 mm, SENSE factor: 3.6, Halfscan factor: 0.61. A whole brain histogram, averaging all subject data, was created. Binary ROI masks were created by thresholding T1 maps using voxel intensities derived from the mean whole brain histogram: white matter (WM): 800-1300ms, grey matter (GM): 1300-1900ms, and cerebrospinal fluid (CSF): 3000ms+). ROI masks were eroded to minimize partial volume effects. The regional mean/standard deviation and median/range of T1 values were calculated. The T1 of venous blood was also considered for a single subject. Here, the sagittal sinus (SS) was manually delineated (fig 3A), and voxels within the SS mask were thresholded based on literature data to include values between 1950-2250ms [6][7]. Average attained PetO₂ values are shown on the x-axis of fig 2B. Neither GM, WM or vessel T1 values appeared affected by the hyper-oxic stimulus (fig 2 A/B). CSF, on the other hand, showed a significant linear decrease as a function of PetO₂. A linear fit to the mean T1 (4275±50, 4206±53, 4172±72 and 4108±63 for baseline, 250mmHg, 350mmHg and 500mmHg, respectively) in CSF yielded the following equation: T1_CSF(ms) = -0.4335*PetO₂+4320 (R2=0.99, RMSE = 5.28). Fitting in terms of the relaxivity (R1=1/T1) as a function of PetO₂ yielded: R1_CSF(ms^-1) = 2.47e^-8*PetO2 + 2.31e^-4 (R2 = 0.99, RMSE 2.90^e-7). Mean T1 values in the sagittal sinus of a single subject increased with PetCO₂ (T1 values shown in fig 3B).Despite the longer T1 relaxation constants at 7T, the qT1 mapping method used in this work did not identify changes in GM/WM tissue T1 in response to hyperoxia. A considerable decrease was observed in CSF, however. The CSF compartment provides a considerable reservoir of unbound water, into which a large amount of plasma O₂ can freely diffuse. Furthermore, large surface arteries come in contact with CSF before penetrating into cerebral tissue. These circumstances may result in a significant reduction in plasma dissolved O₂ before the vasculature penetrates deeper into the cerebrum, thus mitigating an increase in the O₂ gradient between the vasculature and brain tissue. Hyperoxia induces a BOLD effect due to increased venous hemoglobin ([HbV]) [8], which implies that the O₂ gradient driving oxygen diffusion into the tissues is increased. This is supported by the apparent increase in venous T1 (which increases proportional to HbV [7][9]), and suggests that tissue O2 demand is satisfied with an increased proportion of plasma dissolved O₂ rather than O₂ bound to hemoglobin. It remains unclear, however, whether hyperoxia causes a lasting increase in extracellular [O₂] since no T1 effects were observed. It should be noted that in cases of disease, tissue properties are highly variable and local hypoxia may influence the tissue response to elevated O₂ delivery. Hyperoxia did not result in changes in brain tissue T1. It did, however shorten T1 in CSF. Our results challenge the notion that mapping of T1 changes using steady state imaging methods are able to give insight into the actual oxygenation status of tissues. Care should be taken to distinguish actual tissue T1 changes from those which may be related to T2* effects, partial volume effects or increased fluid content such as edema. Furthermore, qT1 mapping is able to distinguish O₂-level dependent changes in venous T1, which may also be used to gain insight into venous hemoglobin saturation.