The discharge patterns of 144 single cortical neurones, within the cutaneous representation of the hand in area 2 (primary somatosensory cortex, SI), were studied in two rhesus monkeys during the performance of an active tactile discrimination task. These were compared to those previously described for units within areas 3b and 1 recorded from the same animals. The task consisted of making a single scanning movement of the digit tips over a surface (first half smooth; second half either smooth or rough). The nature of the texture encountered over the second half of the surface was indicated by the monkey making a differential lever response (push or pull) with the opposite hand. During the task, area 2 units with cutaneous receptive fields (RFs) on the digit tips of interest (those scanned over the surfaces) generally showed an increase in their discharge (75%); patterns of decreased discharge or no modulation (respectively, 12 and 13%) were rarely observed. Units with digital cutaneous RFs not in contact with the stimuli were much more likely to show either a pattern of decreased discharge or no modulation whatsoever (47% in each case), suggesting that there is some selection of cutaneous inputs in this task in that non-active inputs are selectively gated. For units with a cutaneous RF, the sign of modulation changed significantly across SI, in a manner consistent with a pattern of increased convergence onto the more caudal regions of SI. Overall, the proportions of area 2 units with digital RFs on the tips of interest that were classified as either texture-related (25%) or movement-related (26%) were similar to those reported previously for areas 3b and 1, suggesting that their presumed roles in, respectively, the analysis of surface texture and the representation of the physical parameters of movement are shared and distributed across the three cytoarchitectonic subdivisions of SI under consideration. In addition, the discharge patterns of single texture-related cells in areas 3b, 1 and 2 did not reliably signal whether or not the animal successfully discriminated the surfaces, suggesting that information from a population of cells is required for the performance of the task. Texture-related responses in area 2 were, however, unique in two ways. Firstly, 35% of the texture-related units had additional discharges related to the performance of the scanning movement (texture- and movement-related cells); no such units were found in area 3b, and only one was encountered in area 1.(ABSTRACT TRUNCATED AT 400 WORDS)