19F nuclear magnetic resonance spectroscopy was performed at 2.0 Tesla to evaluate the washout of isoflurane from the adult rabbit brain after 90 min of anesthesia. This investigation reconciles previous in vivo NMR studies of others, which observed a slow anesthetic washout, with invasive non-NMR studies that found a rapid washout, as predicted by perfusion-limited models of anesthetic uptake and elimination. Two NMR surface coil experiments were performed: in the first, a 1-cm surface coil was placed directly over the exposed dura to be certain that the washout was observed only from the brain; in the second, a 3-cm coil was placed noninvasively over the intact scalp to emulate previous NMR experiments. As in previous NMR experiments, a slow washout of isoflurane was observed with the large coil. The NMR signal that is observed with the large coil cannot be attributed solely to brain tissue. Fat surrounding the brain contributes significantly to the fluorine NMR spectra that are observed with the 3-cm coil, and its contributions lengthen the apparent washout time. A rapid washout of isoflurane from the rabbit brain was observed with the small coil, whose signal unambiguously arises only from brain tissue. The observed rapid washout is consistent with previous invasive biochemical measurements of anesthetic washout from the brain.