Abstract, Project 1 The overall goal of this proposal is to establish how neural activity drives periarterial CSF pumping and thereby glymphatic clearance of metabolic waste. Project 1 will address that goal via fluid-dynamical modeling of flow at the microscale, flow at the macroscale, and brain-wide clearance – all in both mice and humans. Project 1 will unify the microscale mechanisms and macroscale phenomena measured in Projects 2-4 and deliver predictive, quantitative, testable models. We postulate that neural circuit activity controls glymphatic function at the microscale via dynamics of the neurovascular unit, comprised of an arteriole, the perivascular space (PVS) surrounding it, and the surrounding neuropil. Aim 1 will use detailed fluid-dynamical simulations of the unit, with domain shapes and boundary conditions taken from measurements, and with vasomotion linked empirically to norepinephrine (NE) and acetylcholine (ACh) levels, to characterize and quantify microscale CSF flows and drivers in mice. We postulate that neural activity exerts global control by enlarging and reducing the extracellular space, and through interactions on the network of PVSs. Aim 2 will build a brain-wide hydraulic network model to quantify the effects of global drivers and characterize CSF flow across the entire mouse brain. An essential function of CSF flow in the brain is solute clearance. Aim 3 will build a brain-wide clearance model, taking flows from Aim 2, independently quantifying the effects of advection and diffusion, and accounting for changes in brain state. Aim 4 will build models analogous to those of Aims 1-3, but for humans instead of mice, and supplemented by detailed fluid-dynamical simulations of ventricle flow. This multi-species proposal is designed to reveal how neural circuits control cerebrospinal fluid movement in the mouse and human brain. Project 1 will integrate quantitative measurements of neural activity, blood volume, and CSF movement, from Projects 2-4. The experiments will provide parameters for local and global models, including anatomical shapes, inlet and outlet boundary conditions, and spatiotemporal hemodynamic changes. Models will reveal more information than is accessible experimentally and allow causal manipulations that are impossible in vivo, thereby leading to new hypotheses to be tested. Project 2 will provide PVS shapes and solute efflux measurements (DB53) as well as astrocytic dynamics (via Ca2+ and cAMP sensors). Project 3 will provide spatiotemporal hemodynamic patterns and their dependence on neural and neuromodulatory activity (via Ca2+ and biosensors for NE and ACh). Project 4 will provide data from humans: ventricle and PVS shapes, hemodynamics, and CSF flow in ventricles and PVSs, across spontaneous and sensory-driven neural activity.