Functional magnetic resonance imaging (fMRI) can serve to localize activity in the cerebral cortex. The present study was performed to develop a quantitative means of describing the cortical location activated during voluntary smiling in multiple subjects and to determine whether this location is specific to smiling when compared with other motor tasks. Five human subjects were instructed to smile or to tap the fingers of both hands. Both tasks were performed in a blocked-trial paradigm that consisted of alternating 15-second blocks of a repetitive motor task and 15 seconds of rest. Smiling was also performed as an event-related paradigm in which the subject smiled briefly once every 15 seconds for 20 repetitions that were combined to produce an average response to a single smile. A series of 300 images was acquired using an echo-planar imaging sequence (24-cm field of view; 5-mm slice thickness; repetition time/echo time, 1000/27.2 msec). Each subject's three-dimensional brain images were transformed to Talairach coordinates by stretching or compressing the brain images to fit the standard brain as defined in the Talairach atlas. This allowed data from five subjects to be combined for a numeric description. Functional activation maps acquired by use of the event-related paradigm contained significantly fewer motion artifacts than maps acquired with the blocked-trial paradigm, allowing better visualization of functionally active areas. Three-dimensional Talairach coordinates to describe the locations of peak cortical activity after smiling and finger tapping were established. These coordinates were consistent among subjects. During smiling, statistically significant activation was seen in the motor cortex, primarily along the precentral sulcus; this was inferior and anterior to the region that was associated with finger tapping. This study demonstrates that motion artifacts associated with traditional blocked-trial fMRI protocols can be overcome by employing an event-related paradigm to obtain an average response from a single smile. With the implementation of new imaging paradigms with fMRI, an area of the cerebral cortex has been identified that is specifically activated during voluntary smiling, and remains consistent among subjects. Quantification of fMRI data represents a powerful tool by which to study the cortical response to motor activity and to monitor possible alteration in this activity after injury or surgery. When combined with biofeedback therapy, this technique may help to improve the outcome of facial reanimation procedures in the future.