The authors compared [3H]2-methoxy-idazoxan (RX 821002) and [3H]rauwolscine binding in rat cerebral cortex, spleen and kidney; guinea pig kidney; porcine kidney; human kidney and platelets and HEL and NG 108-15 cells. [3H]RX 821002 had less nonspecific binding and higher affinity than [3H]rauwolscine in most models. Although both ligands detected similar alpha-2 adrenoceptor numbers in rat, porcine and human kidney and in NG 108-15 cells in saturation experiments, [3H]RX 821002 detected more alpha-2 adrenoceptors than [3H]rauwolscine in rat cerebral cortex and spleen, guinea pig kidney, human platelets and HEL cells. These differences were seen in Tris and in glycylglycine buffer regardless of whether EDTA, MgCl2, MgCl2 plus GTP or GTP plus NaCl was added to the former and were not explained by additional labeling of serotonin or dopamine receptors or nonadrenergic sites; in contrast, [3H]rauwolscine also labeled nonadrenergic sites in porcine kidney. In prazosin competition experiments, both ligands differentially recognized alpha-2-adrenoceptor subtypes but this could not account for the observed differences in detected receptor numbers. In epinephrine competition experiments, both ligands labeled similar numbers of agonist low affinity sites in all models; [3H]RX 821002, however, labeled more agonist high-affinity sites than [3H]rauwolscine did in models in which it detected a greater total number of receptors. It was concluded that [3H]RX 821002 is a more suitable ligand for the detection of alpha-2 adrenoceptor than [3H]rauwolscine because of less nonspecific binding, higher affinity and greater specificity for alpha-2 adrenoceptors; moreover, [3H]rauwolscine appears not to detect all agonist high-affinity sites of alpha-2 adrenoceptors.