To evaluate the efficacy of ventricular arrhythmia detection in ambulatory patients with stable coronary artery disease, 101 coronary patients documented by arteriography or remote myocardial infarction underwent concurrent evaluation by 10-hour portable ambulatory ECG monitoring and standard 12-lead electrocardiograms (ECGs) obtained serially over 24 months. Portable ECG recorded premature ventricular contractions (PVCs) in 77 patients: 50% serious PVCs (multifocal, paired, >5/minute, R on T) and 9% patients with ventricular tachycardia. Standard ECG was insensitive (p<.001) in arrhythmia detection both by one tracing (PVC prevalence 17%) obtained within 12 hours of portable ECG or by four serial standard ECGs during the subsequent two weeks (PVC prevalence 18%). Although a total of 1,414 standard ECGs during 24 months recorded PVCs in 49% of patients, serious PVCs were limited (p<.001) to 25% and ventricular tachycardia to 1% of patients. However, ventricular ectopy present by any standard ECG (10 per patient) within three months of ambulatory monitoring was associated with high prevalence of hazardous ectopics by portable ECG (p<.05): serious PVC 92%; ventricular tachycardia 17% patients. Even in patients free of ventricular ectopy by 24-month serial standard ECG, portable ECG detected serious PVCs in 62% and ventricular tachycardia in 6% of patients. Thus stable coronary patients manifested frequent ventricular ectopics by portable ECG which were usually undetected by even multiple serial standard ECGs. Importantly, presence of PVCs by standard ECG was related to increased risk of hazardous ventricular arrhythmias by portable ECG.