Ambulatory ECG had been proposed to examine the amplified high resolution signal-averaged electrocardiogram (SAECG). Clinical investigations are required to confirm the predictive value of such a high resolution technique in arrhythmic risk stratification. The prognostic value of ambulatory Holter SAECG was evaluated in 108 postinfarction patients for the purpose of predicting the occurrence of serious arrhythmic (SARR) events (sudden cardiac death [SCD], VT, or VF) in comparison with classical real-time SAECG. During the 42+/-8 months of follow-up, the sudden cardiac death mortality was 4.6% (five deaths), six (5.6%) patients had VT, and one (0.9%) VF. QRSd was found to be the most predictive parameter using ROC curves analysis for SAAR + outcome (W = 0.833 and W = 0.803 for 25-250 Hz and 40-250 Hz filters, respectively) followed by RMS (W = 0.766 and W = 0.721) and LAS (W = 0.759, W = 0.709) (all P < 0.01). Abnormal Holter SAECG for 25 and 40-Hz LP filter were significant predictors of SARR+ by log-rank test (P < 0.01, P < 0.05, respectively). This study confirms that valuable prognostic information can be obtained from the ambulatory high resolution ECG technique and that Holter SAECG may predict arrhythmic risk in a postinfarction population.