There is an ongoing discussion on whether or not high beta(1)-adrenoceptor selectivity of beta-adrenoceptor antagonists may be favorable in the treatment of patients with heart failure. The present study compared the beta(1)-adrenoceptor selectivity of nebivolol and bisoprolol with that of carvedilol in the human myocardium, using a binding assay in conjunction with either the hydrophilic ligand (+/-)-[3H]4-(3-tertiarybutylamino-2-hydroxypropoxy)-benzimidazole-2-on HCl ([3H]CGP 12.177) or the lipophilic ligand [125I]iodocyanopindolol as radiolabeled compound. Measurements were made using membrane preparations obtained from identical nonfailing donor hearts. beta-adrenoceptor density was found to be slightly higher when [125I]iodocyanopindolol was used compared to [3H]CGP 12.177 (256+/-15 and 213+/-18 fmol/mg protein, respectively). When the highly beta(1)-adrenoceptor-selective compound 2-hydroxy-5-(2-(hydroxy-3-(4((1-methyl-4-trifluoromethyl)-1-H-imidazol-2-yl)-phenoxy)-propyl)-aminoethoxyl)-benzamide (CGP 20.712A) and the highly beta(2)-adrenoceptor-selective compound erythro-(+/-)-1-(7-methylindan-4-yloyl)-3-isopropylaminobutan-2-ol HCl (ICI 118.551) were used in competition experiments, a similar proportion of beta(1)-adrenoceptors was seen for [3H]CGP 12.177 (69.3+/-1.6%) and for [125I]iodocyanopindolol (67.0+/-2.1%). K(i)(beta(1)) and K(i)(beta(2)) were obtained in the presence of 50 nM ICI 118.551 and 300 nM CGP 20.712A. The rank order of beta(1)-adrenoceptor selectivity (K(i)(beta(2))/K(i)(beta(1)) ratio) was nebivolol (for [3H]CGP 12.177 46.1 and for [125I]iodocyanopindolol 22.5)>bisoprolol (13.1 and 6.4)>carvedilol (0.65 and 0.41). To investigate whether in vivo metabolized nebivolol retains high beta(1)-adrenoceptor selectivity, serum specimens were collected before and 2 h after oral administration of 5 mg nebivolol. The samples were used for [125I]iodocyanopindolol binding studies with the myocardial membrane preparations. In these samples, the binding of [125I]iodocyanopindolol to beta(1)-adrenoceptors was inhibited by 46.4+/-5.3%, whereas the binding to beta(2)-adrenoceptors was inhibited by 20.5+/-1.1% compared to that of control samples. It is concluded that nebivolol is approximately 3.5 times more beta(1)-adrenoceptor-selective than bisoprolol in the human myocardium. Furthermore, in vivo metabolized nebivolol retains beta(1)-adrenoceptor selectivity.