The effect on left ventricular function of a gradual withdrawal of chronic metoprolol treatment in postinfarction patients was studied. All patients were in a randomized double-blind post-infarction study with metoprolol (M 100–200 mg daily; N=14) or placebo (P; N =18). After three years treatment the study medication was gradually withdrawn during one week. M-mode echocardiography, guided by concomitant cross-sectional recordings, were performed before, one and 12 weeks after the withdrawal. Treatment (i.e. M or P) had to be reinstituted in eight patients (5 M; 3P) because of the development of disabling symptoms during the follow-up. Heart rate was lower in patients treated with M (57±4) than with P (69±10) (p<0.01). One week after withdrawal of M, heart rate had increased to 77± 13(p<0.001), while patients on P showed no significant change. In order to minimize the influence of heart rate on the evaluation of time intervals in the cardiac cycle, heart rate dependent correction factors were used. One week after M withdrawal there was a prolongation of the pre-ejection period (PEP) from 120±15 ms to 133±16 ms (p< 0.01), mainly due to a prolongation of the interval for early isovolumetric contraction (Q Mc) from 87±10 ms to 101±11 ms (N=11; p≤0.001). Simultaneously, values for isovolumetric relaxation increased from 228±28ms to 286±39 MS (n = 11; p≤0.001), starting from a somewhat lower value than P before withdrawal, reaching an insignificantly higher level and returning to the levels of P. During withdrawal of P stable values were encountered. Twelve weeks after withdrawal, there were no longer significant differences between M and P groups. In conclusion, after a one week gradual withdrawal of M in patients with ischaemic heart disease, a transient increase of both isovolumetric contraction and relaxation phases occur, suggesting depressed myocardial function, despite a transient rebound increase in heart rate.