The antinociceptive effect of subcutaneously (s.c.), intracerebroventricularly (i.c.v.) or intrathecally (i.t.) administered diclofenac was studied in a series of experiments employing the tail-flick (0.01-10.0 mg/kg body weight i.p., 1-50 micrograms i.c.v., 1-10 micrograms i.t.) and hot-plate (0.01-50 mg/kg body weight i.p., 1-50 micrograms i.c.v., 1-10 micrograms i.t.) models representing somatosensory stimuli and the writhing test (0.001 mg-10 mg s.c., 0.1-200 micrograms i.c.v., 0.1-100 micrograms i.t.) and colorectal distension (1-200 micrograms i.c.v.) models representing noxious visceral stimuli. Diclofenac did not exert any antinociceptive effects in the tail-flick or hot-plate models. In the writhing test, diclofenac dose-dependently inhibited the number of writhings after s.c. administration (0.001-10.0 mg/kg body weight) with an ED50 of 1 mg/kg. A similar dose-dependent action of diclofenac was seen after i.c.v. (0.1-200 micrograms) and i.t. (0.1-100 micrograms) administration with an ED50 of 3 micrograms in both cases. The antinociceptive effect of diclofenac administered s.c., i.c.v. or i.t. was almost completely reversed by pretreatment with naloxone, (1 mg/kg body weight s.c.). In the colorectal distension model, the i.c.v. administration of diclofenac (1-200 micrograms), which attenuated the cardiovascular response to colorectal distension, was reversed by naloxone. The pressor and tachycardia response to a 20 s, 80 mmHg colorectal distension was inhibited by diclofenac 100 micrograms i.c.v. (16.1 +/- 1.7 mmHg or 58% and 39.4 +/- 0.4 bpm or 70% versus control, respectively). It is concluded that diclofenac exerts a central, naloxone-reversible antinociceptive action in experimental animals after noxious visceral stimuli but not after somatosensory stimuli.