The antinociceptive effects of the putative endogenous cannabinoid ligand anandamide (ANA) and its fluorinated analog, fluoroanandamide (FA), were determined as measured by the tail-flick and p-phenylquinone (PPQ) stretch tests. The ED50 values (confidence limits) for ANA and FA were 77 (52-13) and 7 (2-21) micrograms/mouse, respectively, for the tail-flick test and 30 (23-41) and 0.5 (0.1-2) micrograms/mouse, respectively, for the PPQ test after intrathecal (i.t.) administration. ANA was not significantly less potent than delta 9-tetrahydrocannabinol (THC) in the tail-flick test, but it was less potent in the PPQ test. FA was more potent than either ANA or THC in tail-flick test. The antinociceptive effects of all drugs (administered i.t.) were blocked significantly or nearly abolished by the pretreatment of the mice with pertussis toxin (i.t.). Pretreatment of the mice with 5 and 25 micrograms forskolin per mouse or 10 micrograms 8-(4-chlorophenyl-thio)-adenosine-3',5'-monophosphate cyclic monosodium salt per mouse (both i.t.) significantly attenuated the antinociception produced by THC but not by ANA or FA. Various calcium modulators were tested in combination with THC, ANA, and FA, but they failed to alter the antinociceptive effects of the drugs. Various potassium channel blockers were tested in combination with the drugs. Apamin, a blocker of small (low)-conductance calcium-gated potassium channels that attenuates THC-induced antinociception, failed to alter ANA- or FA-induced antinociception. In contrast to THC, which is blocked by the kappa antagonist nor-binaltorphimine, ANA- and FA-induced antinociception was not altered by classic opioid antagonists. Also in contrast to THC, which enhances mu and delta opioid-induced antinociceptive effects, ANA failed to significantly alter opioid antinociception. ANA significantly shifted the THC dose-effect curve to the right. Thus, ED50 for DMSO/THC in the tail-flick test was shifted from 14 (7-29) to 54 (38-77) micrograms/mouse and was shifted in the hot-plate test from 22 (12-42) to 63 (43-92) micrograms/mouse. The magnitude of the shift in the ED50 was 3.8-fold in the tail-flick test and 2.9-fold in the hot-plate test. The shifts were parallel and significant. The Ki for the displacement of 3H-CP 55,940 binding by ANA and FA was 214 nM (+/- 45 S.E.M.) and 72 nM (+/- 5 S.E.M.), respectively, in pure spinal cord synaptosomes from the rat. ANA and FA were significantly cross-tolerant to THC. Although similarities between ANA and cannabinoids were shown, several marked differences were observed between ANA and the classic cannabinoids. ANA appears to function as both a cannabimimetic and a blocker of cannabinoid-induced antinociception.