Nociceptive flexion reflexes (RIII) obtained by stimulation of sural nerve were studied in patients with intractable chronic pain before and after functional neurosurgery, either dorsal column stimulation (DCS, n = 15) or posterior selective rhizotomy in the dorsal root entry zone (DREZ, n = 5). Dynamic study of RIII at supraliminal levels provided direct, quantitative and replicable evidence of the inhibition of nociceptive spinal reflexes by DSC. The effects of DSC on the RIII were highly correlated with subjective pain relief. In non-collaborative patients it was still possible to select the best DCS parameters (frequency, intensity) as those providing maximal RIII depression. After posterior selective rhizotomy in the DREZ involving S1-S2 root levels postoperative evidence of selective extralemniscal lesioning could be assessed by the abolition or strong attenuation of nociceptive RIII, whereas preservation of the lemniscal pathways was evidenced by somatosensory evoked potentials. Routine recording of nociceptive reflexes in man proved to be a useful tool for the objective evaluation of anatomo-physiological effects of functional neurosurgical procedures.