Functional MRI (fMRI) is widely used to study brain function in the neurosciences. Unfortunately, conventional fMRI only indirectly assesses neuronal activity via hemodynamic coupling. Diffusion fMRI was proposed as a more direct and accurate fMRI method to detect neuronal activity, yet confirmative findings have proven difficult to obtain. Given that the underlying relation between tissue water diffusion changes and neuronal activity remains unclear, the rationale for using diffusion MRI to monitor neuronal activity has yet to be clearly established. Here, we studied the correlation between water diffusion and neuronal activity in vitro by simultaneous calcium fluorescence imaging and diffusion MR acquisition. We used organotypic cortical cultures from rat brains as a biological model system, in which spontaneous neuronal activity robustly emerges free of hemodynamic and other artifacts. Simultaneous fluorescent calcium images of neuronal activity are then directly correlated with diffusion MR signals now free of confounds typically encountered in vivo. Although a simultaneous increase of diffusion-weighted MR signals was observed together with the prolonged depolarization of neurons induced by pharmacological manipulations (in which cell swelling was demonstrated to play an important role), no evidence was found that diffusion MR signals directly correlate with normal spontaneous neuronal activity. These results suggest that, whereas current diffusion MR methods could monitor pathological conditions such as hyperexcitability, e.g., those seen in epilepsy, they do not appear to be sensitive or specific enough to detect or follow normal neuronal activity.D eveloping a direct MRI method to detect neuronal activity in vivo and noninvasively is a major focus in neuroscience. Progress in this area is required to improve our understanding of normal brain function, and in a clinical setting, to develop new tools for studying normal and abnormal development to diagnose diseases and disorders of the brain. Functional MRI (fMRI) has been widely used in the cognitive neurosciences since its invention in the 1990s (1-3). The most widely used fMRI method, blood-oxygenation-level-dependent (BOLD) MRI, detects hemodynamic changes in the brain, which only indirectly reflects neuronal activity. Moreover, its hemodynamic origin limits both its spatial and temporal resolution and its interpretation as a direct proxy for neuronal activity (4, 5).More recently, several MRI methods were proposed to provide more direct measures of neuronal excitation (6). In particular, diffusion MRI, a method to measure the apparent diffusivity of water within tissues (7-9), has been suggested as a direct functional imaging method to detect neuronal activity (10-13). Early in vivo experiments in both humans and animals reported small but significant increases in highly diffusion-weighted MRI signals, which were ascribed to changes directly induced by the underlying neuronal activity rather than indirect hemodynamic changes (10-13).In vitro experimen...