Drug self-administration methods were used to test the hypothesis that rats would self-medicate with a cannabinoid CB(2) agonist to attenuate a neuropathic pain state. Self-medication of the CB(2) agonist (R,S)-AM1241, but not vehicle, attenuated mechanical hypersensitivity produced by spared nerve injury. Switching rats from (R,S)-AM1241 to vehicle self-administration also decreased lever responding in an extinction paradigm. (R,S)-AM1241 self-administration did not alter paw withdrawal thresholds in sham-operated or naive animals. The percentage of active lever responding was similar in naive groups self-administering vehicle or (R,S)-AM1241. The CB(2) antagonist SR144528 blocked both antiallodynic effects of (R,S)-AM1241 self-medication and the percentage of active lever responding in neuropathic (but not naive) rats. Neuropathic and sham groups exhibited similar percentages of active lever responding for (R,S)-AM1241 on a fixed ratio 1 (FR1) schedule. However, neuropathic animals worked harder than shams to obtain (R,S)-AM1241 when the schedule of reinforcement was increased (to FR6). (R,S)-AM1241 self-medication on FR1, FR3, or FR6 schedules attenuated nerve injury-induced mechanical allodynia. (R,S)-AM1241 (900μg intravenously) failed to produce motor ataxia observed after administration of the mixed CB(1)/CB(2) agonist WIN55,212-2 (0.5mg/kg intravenously). Our results suggest that cannabinoid CB(2) agonists may be exploited to treat neuropathic pain with limited drug abuse liability and central nervous system side effects. These studies validate the use of drug self-administration methods for identifying nonpsychotropic analgesics possessing limited abuse potential. These methods offer potential to elucidate novel analgesics that suppress spontaneous neuropathic pain that is not measured by traditional assessments of evoked pain.