1. Little is known about the effect of central and peripheral nervous system injury on the processing of somatosensory information at the thalamic level in humans. The role of the human thalamic ventrocaudal nucleus (Vc) in nociception is not well understood because reports of nociceptive neuronal responses and stimulation-evoked pain are rare. In this study, we have characterized effects of microstimulation in the tactile region of Vc. Specifically, we investigated the incidence of painful sensations evoked by thalamic microstimulation in patients with and without chronic pain. 2. Data were obtained during stereotactic thalamic procedures for relief of pain or motor disorders. Patients were divided into three groups, those with 1) central poststroke pain (PSP, n = 13); 2) nonstroke pain (NSP, n = 23); and 3) movement disorders (controls, n = 24). Most (15 of 23) of the NSP patients had peripheral nerve damage. Tungsten microelectrodes were used to record neuronal responses in the thalamus and to deliver stimuli. Localization of tactile Vc was determined according to stereotactic coordinates and neuronal responses to innocuous somatic stimuli. At selected sites, microstimulation (1-s trains, 300 Hz, 0.1-0.2 ms pulses, < 100 microA) was performed and the patient was requested to describe the quality of the sensation and its peripheral location (projected field, PF). 3. Microstimulation in tactile Vc commonly evoked paresthesia-type sensations. Threshold stimulation never evoked pain in the NSP patients and evoked pain at only 2% of Vc sites in the movement disorder patients. In these latter 2 groups of patients, stimulation at > 98% of Vc sites evoked paresthesia. By contrast, in the PSP patients, 28% of Vc sites stimulated evoked painful sensations at threshold. Suprathreshold stimuli evoked painful sensations at 46% of Vc sites in the PSP patients but at only 8% of Vc sites in NSP patients and 12% of Vc sites in the movement disorder patients. 4. The thresholds to evoke paresthesia in the NSP and movement disorder patients were significantly lower than the thresholds in the PSP patients. However, stimulation thresholds to elicit pain were similar in all patient groups. 5. All patients were capable of differentiating stimulation-evoked paresthesia from pain. Stimulation-evoked painful sensations in the PSP patients were often described as burning and sometimes as "sharp," "shocking," or "unpleasant." By contrast, the quality of pain evoked in the other patient groups was typically described as unpleasant or shocking. Pain could be evoked at sites throughout tactile Vc, although most sites were located in the ventral 2/3 of the nucleus. 6. In the movement disorder patients, the location of the projected sensation usually corresponded to the location of the receptive fields of the tactile neurons recorded at the same site. By contrast, in both groups of pain patients there was a high incidence of mismatches between the projected and receptive fields. 7. These results suggest that the effective thalamic output from Vc to the cortex is affected by somatosensory deafferentation in pain patients. In addition, in the PSP patients there are also changes in the thalamocortical processing of noxious information. The increased incidence of thalamic-evoked pain in PSP patients may be due to 1) loss of low-threshold mechanoreceptive thalamic neurons such that nociceptive neuronal output is now prominent, 2) reduced tonic inhibition of thalamic or cortical nociceptive neurons, and/or 3) unmasking or strengthening of nociceptive pathways.