Using guinea pig, rat, and human brain membranes depleted of mu and delta receptors by pretreatment with the site-directed acylating agents BIT (mu selective) and FIT (delta selective), previous studies from our laboratory resolved two subtypes of the kappa(2) binding site, termed kappa(2a) and kappa(2b). In more recent studies, we used 6 beta-[(125)Iodo]-3,14-dihydroxy-17-cyclopropyylmethyl-4,5 alpha-epoxymorphinan ([I-125]IOXY) to characterize multiple kappa(2) binding sites in rat brain. The results indicated that [I-125]IOXY, like [H-3]bremazocine, selectively labels kappa(2) binding sites in rat brain membranes pretreated with BIT and FIT. In the rat brain, using 100 nM [D-Ala(2)-MePhe(4),Gly-ol(5)] enkephalin to block [I-125]IOXY binding to the kappa(2b) site, we resolved two subtypes of the kappa(2a) binding site. In the present study we examined the binding of [I-125]IOXY to the kappa(2) receptors of guinea pig brain. As observed in rat brain, [I-125]IOXY, under appropriate assay conditions, selectively labels kappa(2) binding sites. Quantitative binding studies readily demonstrated the presence of kappa(2a) and kappa(2h) binding sites. The kappa(2a) binding sites were selectively assayed using 5 mu M [Leu(5)]enkephalin to block [I-125]IOXY binding to the kappa(2a) sites, and kappa(2b) sites were selectively assayed using 5 mu M (-)-(1S,2S)-U50,488 to block [I-125]IOXY binding to the kappa(2a) sites. Under these conditions, two subtypes of the kappa(2a) site were resolved with high (kappa(2a-1)) and low (kappa(2a-2)) affinity for nor-BNI (K-i values = 0.88 and 476 nM) and CI977 (K-i values = 17.5 and 95,098 nM). Similarly, two subtypes of the kappa(2b) site were observed with high (kappa(2h-1)) and low (kappa(2b-2)) affinity for [D-Ala(2)-MePhe(4),Gly-ol(5)]enkephalin (DAMGO) (K-i values = 97 and 12,321 nM) and alpha-neoendorphin (K-i values = 33 and 5308 nM). Two-site models were also resolved in the presence of 100 mu M 5'-guanylyimidodiphosphate (GppNHp). We carried out detailed ligand selectivity analysis of the multiple kappa(2) binding sites. Most test agents were either nonselective or selective for the kappa(2a-2) site. Nalbuphine was moderately selective for the kappa(2a-2) site. Similarly, although most test agents were either nonselective or selective for the kappa(2b-1) site, butorphanol, and the delta antagonists naltrindole, naltriben, and 7-benzylidene-7-dehydronaltrexone were moderately selective for the kappa(2b-2) site. Of the endogenous opioid peptides tested, BAM22P had the highest affinity for the kappa(2b-2). Site (31 nM) and peptide E had the highest affinity for the kappa(2a-2) binding site (53 nM). These data provide additional evidence for heterogeneity of the kappa-opioid receptor and new targets for drug design, synthesis, and therapeutics.