To examine whether the downregulation of beta-adrenoceptors is accompanied by reduced beta-adrenoceptor-mediated effects in atrial as well as in ventricular myocardium, we investigated the beta-adrenoceptor-effector coupling in atrial and papillary muscle strips from patients with terminal heart failure (heart transplantation because of dilated cardiomyopathy; New York Heart Association Class IV, NYHA IV) and moderate heart failure (mitral valve replacement, NYHA II-III) and in tissue from non-failing hearts. The isometric force of contraction induced by isoprenaline (0.001-1-mu-moll-1) or Ca2+ (1.8-15 mmoll-1) in atrial muscle strips and papillary muscle strips has been measured. We also examined the number of beta-adrenoceptors in both tissues by radioligand binding. The degree of heart failure affected neither the potency (EC50: control: 0.01 (0.001-0.082)-mu-moll-1; NYHA II-III: 0.01 (0.001-0.125)-mu-moll-1; NYHA IV: 0.01 (0.001-0.160)-mu-moll-1) nor the efficacy (NYHA IV: 7.8 +/- 1.0 mN; NYHA II-III: 6.1 +/- 0.7 mN; control: 7.7 +/- 0.9 mN) of the isoprenaline-mediated increase in force of contraction in atrial muscle strips. This is in spite of a reduced number of beta-adrenoceptors in moderately (NYHA II-III) and terminally (NYHA IV) failing atrial myocardium compared to non-failing atrial myocardium (P < 0.05). In contrast, in papillary muscle strips increasing degrees of heart failure were accompanied by a progressive reduction of the isoprenaline-mediated increase in force of contraction (P < 0.05) as well as by a progressive decrease of beta-adrenoceptors (P < 0.05). In papillary muscle strips, there was a significant (P < 0.05) rightward shift of the isoprenaline concentration-response curve in failing myocardium as compared to controls. In addition, the potency of isoprenaline to increase force of contraction was lower in left ventricular myocardium compared to right atrial tissue. The Ca2+-mediated maximal increase in force of contraction in atrial and papillary muscle strips was similar in non-failing and failing myocardial tissue. In conclusion, a surplus of beta-adrenoceptors for mediating positive inotropic effects seems to exist only in atrial but not in left ventricular human myocardium. Therefore, the efficacy of the beta-adrenoceptor-effector coupling, at least in terms of positive inotropic effects, cannot be inferred for any tissue in the human myocardium.