There is compelling evidence indicating that hyperalgesia, allodynia and ongoing pain associated with peripheral nerve injury are due to changes in primary afferent neurons. Nerve injury-induced hyperalgesia and allodynia reflect, at least in part, changes in the excitability (Campbell et al., 1992; Koltzenburg et al., 1994) and/or phenotype (Sato and Perl, 1991; Noguchi et al., 1995) of primary afferent neurons. One particularly important change is the development of ongoing or ectopic activity. Clinical observations indicate that ectopic activity contributes to ongoing neuropathic pain (Campbell et al., 1992; Gracely et al., 1992). Furthermore, ectopic activity may induce changes in the central nervous system that are thought to be responsible for allodynia or light touch-evoked pain (Gracely et al., 1992; Torebjork et al., 1992). At least two subpopulations of primary afferents develop ectopic activity in the presence of nerve injury: injured afferents and their uninjured neighbors. Thus, both populations of afferents are hypothetically capable of initiating as well as maintaining the behavioral changes observed in the presence of nerve injury. Which of these populations is responsible for the initiation and maintenance of neuropathic pain is a question that has been addressed in several recent studies. Results from these studies indicate that this issue has yet to be resolved. Kim and Chung (1992) developed a nerve injury model that has made it possible to determine the relative contribution of injured and uninjured afferents to the behavioral changes observed following injury. In this model, referred to as the spinal nerve ligation (SNL) model of nerve injury, the spinal nerve (ventral ramus) of one or more spinal ganglia is tightly ligated and/or cut at a point before the spinal nerve joins a common nerve. Thus, injured and uninjured afferents are unequivocally identifiable and therefore relatively easy to study in isolation. Sheen and Chung (1993) obtained one of the most compelling pieces of evidence in support of a role of injured afferents in the initiation and maintenance of neuropathicpain'behaviors observed in the presence of a SNL. These authors reported that SNL-inducedpain'behaviors were eliminated and/or prevented by cutting the dorsal roots of the injured spinal nerves (L5 and L6). Yoon et al.(1996) repeated the observations of Sheen and Chung (1993) and extended their observations by investigating the role of uninjured afferents in the behavioral changes observed following SNL. Importantly, cutting the central root of the uninjured ganglia (ie L4) eliminated evoked responses (ie mechanical and cold allodynia), while having no influence on spontaneouspain'behaviors. These observations are consistent with the hypothesis that mechanical and cold allodynia reflect evoked activity in uninjured afferents that impinge on a sensitized central nervous system, and that these SNL-induced behavioral changes are initiated and maintained by activity originating from the injured afferents. Further support for theinjured afferent'hypothesis comes from three studies. Han et al.(2000); CN Liu et al.(2000); X. Liu et al.(2000) have each attempted to correlate the time course of the behavioral changes in response to a SNL injury with that of the ectopic activity arising from the injured nerve. Data from all three groups indicates that ectopic activity develops rapidly over the first 1ą3 days post-injury, and then slowly declines over the subsequent weeks. This pattern of activity is similar to that observed in A-fibers following axotomy of the sciatic nerve at mid-thigh (Devor, 1994). Importantly, this pattern of change in ectopic activity correlates with the pattern of changes in behavior …