Recently, we reported that lysophosphatidic acid (LPA) induces long-lasting mechanical allodynia and thermal hyperalgesia as well as demyelination and upregulation of pain-related proteins through one of its cognate receptors, LPA1. In addition, mice lacking the LPA1 receptor gene (lpa1-/- mice) lost these nerve injury-induced neuropathic pain behaviors and phenomena. However, since lpa1-/- mice did not exhibit any effects on the basal nociceptive threshold, it is possible that nerve injury-induced neuropathic pain and its machineries are initiated by LPA via defined biosynthetic pathways that involve multiple enzymes. Here, we attempted to clarify the involvement of a single synthetic enzyme of LPA known as autotaxin (ATX) in nerve injury-induced neuropathic pain. Wild-type mice with partial sciatic nerve injury showed robust mechanical allodynia starting from day 3 after the nerve injury and persisting for at least 14 days, along with thermal hyperalgesia. On the other hand, heterozygous mutant mice for the autotaxin gene (atx+/-), which have 50% ATX protein and 50% lysophospholipase D activity compared with wild-type mice, showed approximately 50% recovery of nerve injury-induced neuropathic pain. In addition, hypersensitization of myelinated Abeta or Adelta-fiber function following nerve injury was observed in electrical stimuli-induced paw withdrawal tests using a Neurometer. The hyperalgesia was completely abolished in lpa1-/- mice, and reduced by 50% in atx+/- mice. Taken together, these findings suggest that LPA biosynthesis through ATX is the source of LPA for LPA1 receptor-mediated neuropathic pain. Therefore, targeted inhibition of ATX-mediated LPA biosynthesis as well as LPA1 receptor and its downstream pathways may represent a novel way to prevent nerve injury-induced neuropathic pain.