A variety of data support the hypothesis of an opiate receptor complex composed of distinct, yet interacting mu and delta binding sites (termed mu cx and delta cx to indicate binding sites 'in the complex'), in addition to independent mu and delta binding sites, termed mu ncx and delta ncx, to indicate binding sites 'not in the complex'. Ligand binding studies using membranes and slide-mounted sections of rat brain support the hypothesis that the irreversible mu-antagonist beta-funaltrexamine (FNA) selectively alkylates the opiate receptor complex, altering the binding of mu agonists to the mu cx binding site and the binding of [3H][D-Ala2,D-Leu5]enkephalin to the delta cx site. Previous studies demonstrated that the chronic administration of morphine to rats selectively 'upregulates' the opiate receptor complex. In contrast, the chronic administration of naltrexone upregulates several types of opioid receptors, including kappa, the delta ncx binding site, and multiple binding sites labeled by mu agonists. A prediction based upon these observations is that, using [3H][D-Ala2,MePhe4,Gly-ol5]enkephalin to label mu binding sites, chronic morphine should upregulate only the mu cx binding site, whereas chronic naltrexone should additionally up-regulate the mu ncx binding site. In this study we test and confirm this hypothesis, using sensitivity to FNA to define the mu cx binding site. The implications of these data for models of the opioid receptors and the mechanism(s) of tolerance and dependence are discussed.