The positive inotropic effect of milrinone was investigated in isolated, electrically driven (1 Hz) human papillary muscle strips from nonfailing myocardium (control) and from patients with moderate (NYHA II-III) and severe (NYHA IV) heart failure. In the control hearts, milrinone increased force of contraction to about the same degree as Ca2+ at 15 mmol/l. In NYHA II-III, the positive inotropic effect was significantly reduced compared with the control hearts. In NYHA IV, the effectiveness (maximal increase in force of contraction) and potency (as judged from the EC50 values) were significantly less than in NYHA II-III and controls. Preexposure of the preparations to isoprenaline in NYHA II-III and NYHA IV shifted the concentration-response curve for milrinone to the left and, moreover, restored the effectiveness of the compound (i.e., similar positive inotropic effect as Ca2+). No difference in the cAMP-phosphodiesterase inhibition by milrinone could be detected between controls, NYHA II-III, and NYHA IV. It is concluded that diminished basal cAMP production is responsible for reducing the effectiveness of milrinone in the failing human heart. A diminished inhibition of cAMP-phosphodiesterase does not play a role. The fact that stimulation of cardiac beta-adrenoceptors increased the effectiveness of milrinone provides evidence that combined application of catecholamines and phosphodiesterase inhibitors may be useful in the treatment of patients with terminal heart failure.