Abstract, Project 4 During NREM sleep, large waves of cerebrospinal fluid (CSF) flow appear in the brain. Neural slow waves precede these CSF flow waves by several seconds, suggesting that coherent neural activity could drive CSF flow by inducing large-scale neurovascular coupling. This project will test the hypothesis that the spatiotemporal dynamics of neural activity regulate how CSF flows in the human brain. We will optimize and integrate multiple novel MR technologies to image CSF flow in the ventricles and in perivascular spaces, with simultaneous measures of neural activity using EEG, and hemodynamic responses using fMRI. In Aim 1, we will test whether spatiotemporally patterned neural activity is linked to CSF flow, measured during well-established visual tasks. In Aim 2, we will test whether the dynamics of spontaneous neural activity during sleep are coupled to CSF flow at multiple spatial scales. In Aim 3, we will identify how activity in basal forebrain and locus coeruleus, which release norepinephrine and acetylcholine, is linked to the relationship between coherent neural activity and CSF flow. Together, these Aims will test how neural activity is linked to CSF flow across sleep and wakefulness in the human brain, from ventricles down to perivascular spaces. Project 4 will contribute to the overall U19 goals by determining the link between neural activity and CSF flow at multiple spatial scales. By performing whole-brain imaging in humans, we can test how both local and global neural dynamics are linked to CSF flow in the tiny perivascular spaces and in the much-larger ventricles. By interacting with Project 2, our EEG-based measures of neural coherence will be linked to cellular-level measures of neuronal activity, with both our projects testing the link to CSF flow in the perivascular space. Our interactions with Project 3 will be critical to inform how neuromodulators shape vascular dynamics, and how neurovascular coupling is altered across sleep and wakefulness. Finally, our continuous interactions with Project 1 will be essential in order to determine which empirical measurements are needed to fully specify the model, and then update the model to make further predictions.