Background: Results of 24-hour Holter monitoring in elderly patients are often unhelpful, since the prevalence of asymptomatic arrhythmias increases and their prognostic significance is unclear. We investigated the value of the resting electrocardiogram (ECG) in predicting significant findings on 24-hour Holter recordings in those suspected of having cardiac syncope. Objective: To see whether the resting 12-lead ECG can be used as a screening tool to select elderly patients suspected of having cardiac syncope for 24-hour ECG monitoring. Method: Comparison of resting 12-lead ECGs and 24-hour Holter tapes in 145 consecutive elderly outpatients suspected of having a cardiac cause for falls, dizziness, or syncope. Results: Four of 30 normal ECGs (13%) showed an abnormality on Holter monitoring as compared with 55 of the 115 abnormal ECGs (47.8%; chi = 11.7143, p < 0.005). In the ‘normal’ group the 4 abnormal Holter recordings all showed short runs of supraventricular tachycardia, and no intervention resulted. The 115 abnormal resting ECGs showed either ischaemia (n = 27), dysrhythmia (n = 28), sinus bradycardia (n = 22), or conduction defects (n = 38). The 55 of these which showed abnormalities on Holter recordings occurred mostly where the resting ECG showed dysrhythmia (n = 14/28; 50%), bradycardia (n = 19/22; 86.4%), and conduction defect (n = 17/38; 44.7%). Seven patients had complete heart block on Holter, and all had conduction defects on resting ECG (p < 0.0004). Fifteen patients had pauses of longer than 3 s on Holter; all had conduction defect, bradycardia, or dysrhythmias on resting ECG (p < 0.0045). Sixteen patients were paced because of complete heart block or pauses on Holter recordings, and all had either bradycardia or conduction defects on resting ECG, resulting in complete resolution of their symptoms. Conclusions: Patients with suspected cardiac syncope and normal resting ECGs are unlikely to reveal significant abnormalities on single 24-hour Holter monitoring. Cardiac event recorder or prolonged Holter monitoring may be required in patients with strong clinical history. Those with abnormal ECGs, in particular sinus bradycardia and conduction defects, are highly likely to have significant abnormalities on 24-hour ECG monitoring.