1. The present study investigated whether or not there may be differences in the direct cardiac actions of the novel, highly beta(1)-selective adrenoceptor antagonist nebivolol (NEB) in comparison to metoprolol (MET), bisoprolol (BIS), carvedilol (CAR) and bucindolol (BUC) in human myocardium (n=9). 2. The rank order of beta(1)-selectivity as judged by competition experiments to (3)H-CGP 12.1777 in the presence of CGP 207.12 A (300 nmol l(-1), K(i)beta(2)) or ICI 118.551 (50 nmol l(-1), K(i)beta(1)) were NEB(K(i)beta(2)/K(i)beta(1): 40.7) > BIS(15.6) > MET(4.23) > CAR(0.73) > BUC(0.49). 3. The rank order of the negative inotropic potency of the beta-adrenoceptor antagonists measured in left ventricular trabeculae (dilated cardiomyopathy, DCM) as judged by the concentration needed to induce a 50% decrease in isoprenaline (1 micromol l(-1))-stimulated force (IC(50)) was: MET (0.6 micromol l(-1)) > CAR (4.1 micromol l(-1)) > NEB (7.0 micromol l(-1)). 4. NEB, BUC, MET and CAR did not not exert an intrinsic sympathomimetic activity (ISA) as determined by measurements of force development in forskolin (0.3 micromol l(-1)) pre-treated left ventricular trabeculae, nor by measuring adenylate cyclase activity in forskolin (0.3 micromol l(-1))-stimulated assays (crude membranes). This also holds true for radioligand binding assays with or without guanine nucleotide guanyl-5'-yl imidodiphosphate (Gpp(NH)p). 5. Although all studied beta-adrenoceptor antagonists lack intrinsic sympathomimetic activity (ISA), they differ in the beta(1)-selectivity as well as in their direct negative inotropic action. These differences as well as the mode of extracardiac action may have an impact on outcome of patients treated with beta-adrenoceptor antagonists.