Nicotine was found to be potent in producing antinociception in mice and rats as measured by tail-flick latency but its action was of short duration. Nicotine's ED50 values (confidence limits) were 0.7 (0.4-1.1) and 2.0 (1.2-3.4) mg/kg in rats and mice, respectively, at the time of maximum effect. Brain levels of nicotine reached a maximum at 10 min, whereas antinociception was maximal within 2 min in rats. However, there was a good correlation between the time courses of antinociception and brain levels of nicotine in mice with both attaining maximal levels at 5 min. It was found that tachyphylaxis developed to the antinociception in rats within 10 min and lasted for up to 14 h, but tachyphylaxis did not develop to nicotine-induced antinociception in mice. The effect of nicotine appeared to be central in both rats and mice inasmuch as mecamylamine antagonized completely but hexamethonium (5 mg/kg) antagonized partially. Nicotine-induced antinociception was not blocked in either rats or mice by atropine in doses up to 10 mg/kg. Naloxone did not block nicotine in rats but did antagonize the antinociception in mice. In addition, yohimbine, a selective alpha-2 antagonist, blocked the antinociception in both species. These data implicate several different mechanisms in the antinociceptive action of nicotine.