Naloxone is known to decrease, increase or have no effect on nociceptive thresholds. Here, using two commonly accepted pain-related behaviors (licking and flinching) associated with injection of noxious formalin into a hind paw in rats, naloxone (0.1-1 mg/kg s.c.) simultaneously decreases and increases nociceptive responding in the same animal. Licking, which is reduced by naloxone, is enhanced by low doses but attenuated by high doses of morphine. However, although licking initially increases with a rise in formalin concentration, at higher concentrations the time spent licking the injected paw actually declines. By contrast, flinching, which is enhanced by naloxone, is only antagonized by morphine and increases linearly with formalin concentration. Both actions of naloxone can be interpreted in terms of a leftward shift in the formalin concentration-response curves. This study demonstrates that naloxone can increase formalin-induced flinching while simultaneously decreasing licking behavior. These findings suggest that, on its own, an unexpected decrease in a single nociceptive index may be an inadequate criterion for demonstrating antinociception.