1. To assess the mechanisms used by cortical neurons to sense motion across the skin, we applied pulsatile stimuli to a series of adjacent positions on the glabrous skin of the hand using a computer-controlled OPTACON stimulator. We describe responses of 129 single neurons in primary somatosensory cortex of alert monkeys to a horizontal bar pattern that was displaced proximally or distally in 1.2-mm steps at 10-, 20-, and 40-ms intervals (100, 50, and 25 Hz, respectively). These frequencies span the range in which apparent motion is transformed perceptually in humans from a smooth uninterrupted sweep into a series of distinct pulses that are resolved as separate events. The experiments are thus designed to decipher the neural correlates distinguishing continuous motion from discrete taps. 2. Cortical receptive fields mapped with moving bar patterns spanned 5-24 rows on the tactile array (16.2 +/- 5.4, mean +/- SD). Over 40% of the fields encompassed 18 or more rows (greater than or equal to 21.6 mm), allowing these neurons to integrate spatial information from an entire image displayed on the OPTACON. Cortical receptive fields are considerably larger than those of mechanoreceptors mapped with the same moving bar patterns (4.2 +/- 2.3 rows, mean +/- SD), reflecting convergent inputs in subcortical and cortical relays. Responses were either relatively uniform across the field or strongest at the initial point of entry, depending on the frequency of stimulation. A sharply defined field center was absent from most of the cells recorded in this study. 3. Temporal frequency of stimulation appears to be a major determinant of cortical firing patterns. Low-frequency stimuli are more effective in activating cortical neurons, producing more spikes per sweep and greater phase-locking to individual stimuli than do high frequencies. The total spike output of cortical neurons is proportional to the pulse interval over the range 10-40 ms, increasing linearly by an average of 5.9 spikes/10-ms increase in pulse period. Peak firing rates and modulation amplitude are also highest when pulses are presented at long intervals, falling significantly as the stimulation frequency rises. The reduction in firing at high pulse rates is apparently due to central mechanisms, as both rapidly adapting and Pacinian corpuscle afferents display nearly constant spike outputs and uniform sensitivity within the field when tested with identical bar patterns. Central networks thus behave as low-pass filters, reducing cortical responses to rapidly applied sequential stimuli.(ABSTRACT TRUNCATED AT 400 WORDS)