In the present study, functional magnetic resonance imaging (fMRI) was used to examine pain perception in humans. Three types of noxious stimuli were presented: electric shock (20.8 mA, 2 Hz), heat (48 degrees C), and mechanical, as well as a control tactile stimulus. The significance of activation at the level of the voxel was determined using correlation analysis. Significant region of interest (ROI) activation was determined by comparing the percentage of active voxels in each ROI to activation in a control ROI in the visual cortex. In response to tactile and shock stimuli, consistent activation was seen in the postcentral gyrus, parietal operculum, and ipsilateral cerebellar cortex. No significant cortical activation was detected in response to noxious heat or mechanical stimulation when compared to nonpainful intensity levels. The data did not indicate adaptation, although further study in this area is necessary. Stationary noxious thermal and mechanical stimulation are "pure" noxious stimuli, while electrical stimulation influenced nociceptive and nonnociceptive receptors. Lack of detectable activation in response to pure noxious stimuli supports the idea that nociceptive and nonnociceptive fibers are interspersed in the somatosensory cortex. Conflicting results from recent functional imaging studies of pain perception regarding cortical activation indicate that it is essential to consider both the tactile and nociceptive components of the stimuli used, the spatial extent of stimulation, and the possibility of adaptation to the response. Furthermore, these results suggest that subtractive or correlative methods may not be sufficiently sensitive to image the activity of nociceptive cells, which are sparsely distributed throughout the somatosensory cortex.