The goal of this study was to determine whether opioids of varying potencies are pharmacologically active via the inhalation route of administration in mice. The opioids evaluated included heroin, morphine, codeine, fentanyl and meperidine; each of these drugs has the potential for abuse in humans. Inhalation exposure to each of these compounds produced antinociception in a dose-dependent manner as assessed in the tall-flick test. No pyrolysis products were detected after heating either morphine or codeine at 250 degrees C for 5 min. Although 6-acetylmorphine was found after heating heroin, it accounted for less than 5% of the recovered sample. Heroin was somewhat less potent by inhalation administration than i.v. injection, with ED50 values of 1.6 and 0.69 mumol/kg, respectively. In contrast, the relative potency of morphine was substantially greater when inhaled than when injected, with respective ED50 values of 0.77 and 3.9 mumol/kg. Whereas the body to brain ratios of [3H]morphine were approximately 8 and 20 for inhalation exposure and i.v. injection, respectively, the ratio for heroin was approximately 5 regardless of administration route. This pattern of results suggests that the increase in morphine potency upon inhalation may have resulted from an increased accessibility to the brain compared with i.v. injection. Finally, naloxone reversed the antinociceptive effects of volatilized heroin, but neither the kappa selective antagonist nor-binaltorphimine nor the delta selective antagonist naltrindole blocked this antinociception, which suggests the involvement of mu opioid receptors. These findings taken together suggest the potential for the abuse of a variety of opioids, in addition to heroin, through the inhalation route of administration by humans.