Various hypotheses have been proposed to account for the mechanical hyperalgesia and spontaneous pain seen in animal models of peripheral neuropathy. The purpose of the present study was to determine whether there exists a spinal neuronal correlate to these properties. An experimental neuropathy was induced in male Sprague-Dawley rats by placing a 2-mm PE-90 polyethylene cuff around the sciatic nerve. All rats were subsequently confirmed to exhibit mechanical allodynia in the von Frey test. After induction of anaesthesia with pentobarbital and acute spinalization at T9, electrophysiological experiments were performed, recording extracellular single unit activity from ipsi- and contralateral wide dynamic range dorsal horn neurons in spinal segments L1-4. On-going activity was greater in short-term (11-22 days after cuff implantation) and long-term (42-52 days) cuff-implanted rats; 38 spikes/s in short-term versus 19 spikes/s in controls; 29 spikes/s in long-term ipsi- and contralateral neurons. Receptive fields in controls were always restricted, but in almost all cuff-implanted rats extended over the whole hind paw. Responses to noxious mechanical (pinch) and noxious heat stimulation of the cutaneous receptive field in controls consisted of the typical fast initial discharge followed by an afterdischarge. In all neurons from cuff-implanted rats the initial discharge resembled that in controls. However, the afterdischarge, particularly that in response to noxious pinch, was markedly greater in both magnitude and duration. It is suggested that the greater on-going discharge is the cellular correlate of spontaneous pain, and the potentiation of the afterdischarge in response to noxious stimulation is the correlate of hyperalgesia. Given that acutely spinalized rats were tested, only peripheral and/or spinal mechanisms can be considered to explain these data. Considering all the data, it can be concluded that there is a greater change in fibres mediating noxious mechanical than noxious thermal inputs. Among different hypotheses, the one with which the present data are most compatible is that which proposes that chronic nerve injury or inflammation induces phenotypic changes predominantly in myelinated afferents. There may be a redistribution of membrane-bound ion channels, predominantly sodium channels, which leads to ectopic activity and thus spontaneous discharge of dorsal horn neurons. With regard to mechanical stimulation-evoked synaptic input, the central terminals of myelinated afferents expand into regions of the spinal cord which normally receive their predominant input from unmyelinated nociceptive afferents. This may be coupled with a change in these myelinated afferents so that they now synthesize and release peptides, primarily substance P, from their central terminals with the result that the effects of their chemical mediators of synaptic transmission add to the effects of nociceptive inputs leading to exaggerated responses to painful stimuli, thus the basis of clinical hyperalgesia.