The development of chronically painful states following peripheral nerve injury may involve different mechanisms depending on the nature and extent of the nerve lesion. The altered spinal neurochemistry of two substances, the excitatory amino acid glutamate operating via the N-methyl-D-aspartate receptor and the endogenous opioid peptide dynorphin, have been implicated in behavioral sequelae that follow partial peripheral nerve injury. In addition, dynorphin has nonopioid functions which may involve the N-methyl-D-aspartate receptor. We investigated two hypotheses: that the development of mechanical allodynia following complete nerve injury is not greatly influenced by the N-methyl-D-aspartate receptor, and that spinal dynorphin and glutamate expression is interdependent. These studies employed sciatic cryoneurolysis, a complete but transient peripheral nerve injury that results in a delayed mechanical allodynia beginning 21-28 days after injury. Rats were administered dizocipline maleate (MK-801) at 0.25 mg/kg twice per day intraperitoneally from days 0-7 or from days 0-21 post-lesion to pre-emptively block the N-methyl-D-aspartate receptor. In a separate group of rats, an antibody to dynorphin was administered intraperitoneally at 16.6 mg/kg twice per day from days 14 to 21 post-lesion. For all groups, the outcome of allodynia behavior was assessed using von Frey filaments at 42 days post-lesion and the resulting dynorphin and glutamate immunoreactivity in the substantia gelatinosa was measured using proportional area stained and relative optical density, respectively. Only the 0-7 day MK-801 treatment increased the resulting mechanical thresholds significantly (mean +/- S.E.M. 7.0 +/- 1.2 g) when compared to saline-injected animals (3.9 +/- 0.6 g). However, this effect did not prevent allodynia since baseline thresholds were 12 or 15 g for each group. With regard to resulting spinal immunoreactivity, anti-dynorphin antibody treatment significantly increased glutamate immunoreactivity when compared to saline-treated animals (mean relative optical density +/- S.E.M. = 807.2 +/- 3.6 versus 779.6 +/- 8.3, respectively; P = 0.01) at 42 days post-lesion. We conclude that the development of allodynia following sciatic cryoneurolysis peripheral nerve injury involved a minimal contribution from N-methyl-D-aspartate receptor activity. In addition, this study demonstrated that decreasing available dynorphin using antiserum had a significant and lasting effect on spinal glutamate expression without altering the outcome of allodynia. These conclusions suggest that etiological mechanisms leading to pain behaviors are not equal for all nerve injuries, and that altering kappa opioid levels can affect glutaminergic pathways at a substantially later time.