We have previously reported that scrotonin concentration was reduced in the brain of mice with neuropathic pain and that it may be related to reduction of morphine analgesic effects. To further prove this pharmacological action, we applied fluoxetine, a selective serotonin reuptake inhibitor, to determine whether it suppressed neuropathic pain and examined how its different administration routes would affect antinociceptive and antiallodynic effects of morphine in diabetic (DM) and sciatic nerve ligation (SL) mice, as models of neuropathic pain. Antiallodynia and antinociceptive effect of drugs were measured by using von Frey filament and tail pinch tests, respectively. Fluoxetine given alone, intracerebroventicularly (i.c.v., 15 mu g/mouse) or intraperitoneally (i.p., 5 and 10 mg/kg) did not produce any effect in either model. However, fluoxetine given i.p. enhanced both antiallodynic and antinociceptive effects of morphine. Administration of fluoxetine i.c.v., slightly enhanced only the antiallodynic effect of morphine in SL mice. Ketanserine, a serotonin 2A receptor antagonist (i.p., I mg/kg) and naloxone, an opioid receptor antagonist (i.p., 3 mg/kg), blocked the combined antinociceptive effect of fluoxetine and morphine. Our data show that fluoxetine itself lacks antinociceptive properties in the two neuropathy models, but it enhances the analgesic effect of morphine in the periphery and suggests that co-administration of morphine with fluoxetine may have therapeutic potential in treatment of neuropathic pain.