The noninvasive assessment of patients who present with syncope is based on a thorough, complete history and physical examination. The history requires close attention to precipitating events and the description of the spell. Often, patients are poor historians with regard to symptoms leading up to and following the episode of syncope. Therefore, it is important to interview individuals who observed the spell to improve the accuracy of the history and sort out whether or not the spell was due to cardiogenic syncope, vasodepressor syncope, or a neurologic disorder. Carotid sinus massage is a useful procedure that can be performed during the routine physical examination, identifying patients who are at increased risk for carotid sinus syncope because of hypersensitivity of the carotid sinus. The clinician must be careful to attribute the clinical syncope to carotid sinus hypersensitivity only when the spell induced at the time of carotid massage reproduces the clinical spell. Routine laboratory tests are commonly performed, although the blood tests rarely yield information to confirm the cause of syncope. The routine ECG is often helpful identifying abnormalities of rhythm, conduction, or morphology that give a clue as to the cause for the patient's syncope. The most helpful aspect of ECG recording is to obtain a recording during an episode of syncope when exact correlation can be made between the ECG findings and the patient's symptoms. Recording the ECG during the spell can be achieved using 24-hour ambulatory monitoring, an event recorder, or a memory loop recorder. Twenty-four-hour ambulatory monitoring is useful in patients who have frequent spells that would be expected to be recorded during 1 or 2 days of monitoring. These individuals need to have a non-life-threatening spell and therefore be safe to evaluate as an outpatient. The event recorder and loop memory recorder have proved extremely helpful in evaluating spells that occur too infrequently to be recorded by 24-hour or 48-hour ambulatory monitoring. The nature of these recording devices does require that the patient or a companion be able to activate the monitor at the time of symptoms. If a patient experiences syncope but is unable to activate the device, the important information as to what the rhythm was doing at the time of symptoms is lost. The implantable loop recorder should prove to be uniquely advantageous by allowing extended ECG recording with the device activated by the patient or companion recording 20 minutes before and 4 minutes after device activation. Signal-averaged electrocardiography is most helpful in assessing patients with ischemic heart disease with a substrate capable of supporting a reentrant ventricular arrhythmia. This test should be used in combination with other historical or laboratory predictors of arrhythmic events such as history of myocardial infarction or abnormal ventricular function assessed by echocardiography. In this setting, the signal-averaged ECG helps to identify patients at increased risk for ventricular tachycardia as the cause of syncope who thus may benefit from electrophysiologic testing.