Spinal cord stimulation (SCS) is a treatment given to patients with drug-resistant neuropathic pain, in particular pain resulting from peripheral nerve injury. However, the reasons why some patients develop neuropathic pain and why SCS is not effective in all patients with this chronic pain are not fully understood. The present study compares the response to SCS and the yield of neuropathic animals in variants of the spared nerve injury (SNI) model introduced by Decosterd and Woolf (I. Decosterd, C.J. Woolf, Spared nerve injury: an animal model of persistent peripheral neuropathic pain, Pain 87 (2000) 149-158). Sprague-Dawley rats were prepared with various types of lesions of different branches of the sciatic nerve and then tested for paw mechanical hypersensitivity. A miniature electrode system for SCS was implanted at the T10-T11 vertebral level. Stimulation was applied in awake, freely moving animals with parameters comparable to those employed clinically. Suppression of paw hypersensitivity was considered a positive response to SCS. The incidence of mechanical hypersensitivity ("allodynia") in the different models was: SNI 53%; peroneal axotomy 45%; tibial axotomy 68%; tibial tight ligation 73% and partial tibial tight ligation 50%. "Mirror phenomena" with contralateral paw hypersensitivity was present in about 20% of the animals. The response to SCS differed between models with the lowest response rate in the original SNI model (8%) while the others demonstrated rates in the order of 40-50%. There was a tendency that the efficacy of SCS in suppressing allodynia was inversely related to the severity of hypersensitivity. In conclusion, modifications of the SNI model provide a reproducible incidence of neuropathic hypersensitivity and an increased responsiveness to SCS. These variants may prove suitable for future research on the mechanisms involved in pain relief with SCS.