The explanation for the co-existence of opioid and nonopioid components of tramadol-induced antinociception appears to be related to the different, but complementary and interactive, pharmacologies of its enantiomers. The (+) enantiomer had Ki values of only 1.33, 62.4 and 54.0 microM at mu, delta and kappa receptors, respectively. The (-) enantiomer had even lower affinity at the mu and delta sites (Ki = 24.8, 213 and 53.5 microM, respectively. The (+) enantiomer was the most potent inhibitor of serotonin uptake (Ki = 0.53 microM) and the (-) enantiomer was the most potent inhibitor of norepinephrine uptake (Ki = 0.43 microM). Basal serotonin release was preferentially enhanced by the (+) enantiomer and stimulation-evoked norepinephrine release was preferentially enhanced by the (-) enantiomer. The (+) and (-) enantiomers each independently produced centrally mediated antinociception in the acetylcholine-induced abdominal constriction test (ED50 = 14.1 and 35.0 micrograms i.t., respectively). Racemic tramadol was significantly more potent (P < .05) than the theoretical additive effect of the enantiomers (antinociceptive synergy). Synergy was also demonstrated (P < .1) in the mouse 55 degrees C hot-plate test (i.p. route) and (P < .05) the rat Randall-Selitto yeast-induced inflammatory nociception model (i.v. and i.p. routes). Critically, the enantiomers interacted less than synergistically in two side-effects of inhibition of colonic propulsive motility and impairment of rotarod performance. The racemate and the (+) enantiomer were active in a chronic (arthritic) inflammatory pain model. Taken together, these findings provide a rational explanation for the coexistence of dual components to tramadol-induced antinociception and might form the basis for understanding its clinical profile.