The definition of chronic heart failure (CHF) in clinical practice concentrates on two essential features, namely symptoms suggesting CHF and the objective demonstration of a sufficient severity of cardiac dysfunction. The cardiac dysfunction may be endocardial, myocardial, pericardial, valvular or arrhythmic in nature.Relief of symptoms, at least by the use of diuretic therapy, supports the relationship between cardiac dysfunction and symptoms. However, asymptomatic left ventricular (LV) dysfunction is also common, is associated with a high morbidity and mortality and is amenable to treatment. Therefore it is appropriate to try and identify asymptomatic LV dysfunction.Estimates of the prevalence of CHF in the adult population range from 4 to 20 per 1000 people, rising with age up to 170 per 1000 people aged 70 years and over. Previous estimates of a prevalence of CHF of about 1% (10 per 1000) are broadly supported by more recent data. Estimates of the prevalence of asymptomatic LV dysfunction range from 8 to 59 per 1000, depending on age, the echocardiographic method for measuring ventricular dysfunction and the threshold value of ejection fraction used to separate those with and without LV dysfunction. Asymptomatic LV dysfunction is at least as common and possibly twice as common as CHF.The first step in screening for CHF and LV dysfunction is the recognition of symptoms of CHF and knowledge of the past medical history, in particular a history of myocardial infarction. However, many patients with LV dysfunction will be asymptomatic and in at least 50% of patients with some of the clinical features of CHF the diagnosis cannot be sustained by more detailed investigation.Echocardiography is currently the tool of choice to investigate LV dysfunction although important problems of interpretation of many echocardiographic measurements exist. Echocardiographic examination should be performed on all patients suspected of having CHF before the diagnosis is accepted. It does not generally appear an effective use of resources to screen the general population for asymptomatic LV dysfunction echocardiographically. Screening patients who have had a myocardial infarction (recent or remote) increases the yield of LV dysfunction, with up to 40% of patients surviving hospital admission having important ventricular dysfunction. It is less clear if screening patients with longstanding hypertension or diabetes mellitus is appropriate.Access to echocardiography is restricted in some communities. In these cases, a normal ECG may identify patients at low risk of LV dysfunction. A normal chest x-ray is probably less effective than the ECG in excluding LV dysfunction. However, it should be recognised that it is probably often simpler, more accurate and less expensive to perform echocardiography as the initial investigation.More recently, the potential for natriuretic peptides to identify LV dysfunction has been investigated. Currently, there is sufficient evidence to indicate that these biochemical measures are a useful way of confirming LV dysfunction or CHF in cohorts of patients. Whether natriuretic peptides will prove sufficiently accurate to be used to exclude CHF in individuals remains to be determined. Echocardiography will still be required in those with elevated levels of natriuretic peptides to identify the cause of the cardiac dysfunction.