Our study was designed to study the modulatory role of nitric oxide/cyclic guanosine monophosphate in the spinal cord on the antinociception induced by morphine and beta-endorphin given supraspinally. The antinociception was assessed by the tail-flick test in male ICR mice. The antinociception induced by intracerebroventricularly (i.c.v.) administered morphine was potentiated by intrathecal (i.t) injection of N omega-nitro-L-arginine (2 micrograms), hemoglobin (120 micrograms) or methylene blue (5 micrograms), but was attenuated by i.t. administered L-arginine (20 micrograms) or 3-morpholino-sydnonimine (SIN-1, 5 micrograms). In contrast, the antinociception induced by i.c.v. administered beta-endorphin was attenuated by i.t. pretreatment with N omega-nitro-L-arginine (2 micrograms) and the attenuation of beta-endorphin-induced antinociception by N omega-nitro-L-arginine was reversed by i.t. administered L-arginine (20 micrograms). The antinociception induced by i.c.v. administered beta-endorphin was also attenuated by i.t. administration of hemoglobin (120 micrograms) or methylene blue (5 micrograms). Intrathecal pretreatment with L-arginine did not affect i.c.v. administered beta-endorphin-induced antinociception. It is concluded that the inhibition of nitric oxide/cyclic guanosine monophosphate system in the spinal cord potentiates i.c.v. administered morphine-induced antinociception but attenuates i.c.v. beta-endorphin-induced antinociception. The activation of nitric oxide/cyclic guanosine monophosphate system in the spinal cord attenuates i.c.v. administered morphine-induced antinociception but does not affect i.c.v. administered beta-endorphin-induced antinociception.