This prospective study examined the influence of long-term amiodarone therapy on the parameters of the signal-averaged ECG and their relation to simultaneously derived Holter monitoring data. For this purpose, 23 patients with angiographically confirmed dilated cardiomyopathy or coronary heart disease and high-grade ventricular arrhythmias, in whom an average of four class I antiarrhythmic drugs had proven ineffective, were stabilized on amiodarone. Before the beginning of therapy, as well as after 2 months and, subsequently, every 3 months, a resting ECG, a signal-averaged ECG by Simson's method, and Holter monitoring were performed. Compared to the initial measurement, we found a significant increase in the duration of the total filtered QRS complex from an average of 114 +/- 24 ms to 127 +/- 35 ms, while the change in voltage did not reach the significance level. The incidence of late potentials remained largely constant under amiodarone; 10 patients showed a constant late potential, 12 patients had no late potential, and one patient with coronary heart disease developed a new late potential. In the long-term follow-up, we ascertained a relatively high responder rate under amiodarone between 41% and 81%. No relation could be detected between the results of the signal-averaged ECG and those of 24-h Holter monitoring.