To understand the neurobiology underlying autism spectrum disorders, more data at the cellular level are needed. This is difficult to obtain from living individuals. However, physiological data can be collected directly from cells in the human brain (i.e., neurons) during neurosurgery for medically refractory epilepsy. Two methods can be used in these cases: depth electrode recording and surface grid monitoring. Ralph Adolphs and his colleagues at the California Institute of Technology are planning a project, involving several hospitals, in which these data can be recorded in people with autism who are having surgery for epilepsy. This would represent the first study that involves direct recording from brain cells in people with autism. The data could then be compared with those obtained through functional magnetic resonance imaging, and perhaps eventually with molecular and genetic data. Adolphs’ team plans to begin the experiments by characterizing the basic electrophysiological properties of neurons, such as interspike intervals, spontaneous firing rates and spike waveforms. Next, they aim to present a standardized series of stimuli to all participants across the hospitals to measure evoked responses to a range of stimuli, including faces, landscapes and objects of special interest. This would serve to survey the overall response profiles of neurons in the brains of people with autism. Finally, the investigators plan to examine in detail the neural responses to faces, which are processed abnormally by people with the disorder.