Clinical studies conducted with carvedilol suggest that beta-adrenoceptor antagonism is an effective therapeutic approach to the treatment of heart failure. However, many beta-adrenoceptor antagonists are weak partial agonists and possess significant intrinsic sympathomimetic activity (ISA), which may be problematic in the treatment of heart failure. In the present study, the ISAs of bucindolol, xamoterol, bisoprolol, and carvedilol were evaluated and compared in normal rats [Sprague-Dawley (SD)], in rats with confirmed heart failure [spontaneously hypertensive heart failure (SHHF)], and in isolated neonatal rat cardiomyocytes. At equieffective beta1-adrenolytic doses, the administration of xamoterol and bucindolol produced a prolonged, equieffective, and dose-related increase in heart rate in both pithed SD rats (ED50 = 5 and 40 microgram/kg, respectively) and SHHF rats (ED50 = 6 and 30 microgram/kg, respectively). The maximum effect of both compounds in SHHF rats was approximately 50% of that observed in SD rats. In contrast, carvedilol and bisoprolol had no significant effect on resting heart rate in the pithed SD or SHHF rat. The maximum increase in heart rate elicited by xamoterol and bucindolol was inhibited by treatment with propranolol, carvedilol, and betaxolol (beta1-adrenoceptor antagonist) but not by ICI 118551 (beta2-adrenoceptor antagonist) in neonatal rat. When the beta-adrenoceptor-mediated cAMP response was examined in cardiomyocytes, an identical partial agonist/antagonist response profile was observed for all compounds, demonstrating a strong correlation with the in vivo results. In contrast, GTP-sensitive ligand binding and tissue adenylate cyclase activity were not sensitive methods for detecting beta-adrenoceptor partial agonist activity in the heart. In summary, xamoterol and bucindolol, but not carvedilol and bisoprolol, exhibited direct beta1-adrenoceptor-mediated ISA in normal and heart failure rats.