The residence of morphine in the systemic circulation is prolonged despite a high systemic clearance, suggestive of significant extravascular sequestration. The present study was conducted to test the hypothesis that morphine binds significantly in tissues, and that the liver plays an important role in morphine binding. [14C]Morphine was administered to male Sprague-Dawley rats 55 min before unlabeled morphine or saline. Blood 14C increased immediately after injection of unlabeled morphine; the area under the blood concentration-time curve (AUC) for 14C increased approximately 2-fold after morphine compared with saline injection. Residual radioactivity in the liver was lower in morphine-treated rats than in controls, suggesting that unlabeled drug displaced [14C]morphine (or a metabolite) from binding sites. To examine this phenomenon more directly, a recirculating isolated perfused liver system was employed. [14C]Morphine was added to the perfusate reservoir 15 min before unlabeled morphine or saline; perfusate and bile samples were collected for 120 min. Upon termination of perfusion, the liver was fractionated to identify the hepatic subcellular fraction(s) in which morphine was sequestered. The perfusate AUC for [14C]morphine was increased approximately 2-fold in response to unlabeled drug, consistent with the in vivo experiment. Morphine was associated preferentially with the cytosolic fraction, and [14C]morphine in all relevant fractions was reduced after administration of unlabeled morphine. In contrast, unlabeled drug had no influence on derived [14C]morphine-3-beta,D-glucuronide. These data are consistent with significant, reversible binding of morphine in hepatic tissue.