Marked neuro-endocrine activation in patients with heart failure indicates a worse prognosis and a greater prognostic benefit from the use of ACE inhibitors. However, although the incidence of heart failure rises rapidly with age, relatively little is known about activation of the renin-angiotensin and sympathetic nervous system in patients with heart failure over the age of 75 years. This study was undertaken to investigate plasma concentrations of neurohormonal variables in elderly patients referred to the cardiac clinic with a presumptive, but unconfirmed, diagnosis of heart failure, and to compare these values to plasma concentrations found in age-matched normal subjects. Fifty patients referred with a diagnosis of heart failure were studied. All were receiving a diuretic but not an ACE inhibitor. Patients with renal, haematological and valve disease were excluded. Routine biochemistry and neurohormonal measurements were performed at their first visit, together with an electrocardiogram, chest X-ray and a full clinical examination by an experienced cardiologist. An echocardiogram and Doppler study was also performed and the diagnosis of heart failure either confirmed or refuted. Plasma concentrations of neuro-endocrine variables in healthy elderly subjects were similar to our normal laboratory range in younger subjects with the exception of atrial natriuretic peptide (ANP) (40 +/- 6 pg.ml-1, normal range < 40) and noradrenaline (5.7 +/- 0.7 nmol.l-1, normal range < 2.8). Impairment of left ventricular systolic function was confirmed in 38 of the 50 symptomatic patients (76%) and was associated with increases in plasma concentrations of active renin (58 +/- 8 IU.mol-1, P < 0.001 compared to healthy elderly subjects), angiotensin II (23 +/- 5 pg.ml-1, P < 0.008), noradrenaline (7.7 +/- 1.2 nmol.l-1, P < 0.01) and atrial natriuretic peptide (121 +/- 18 pg.ml-1, P < 0.002). Plasma concentrations were similar in normal subjects and those receiving treatment for heart failure but in whom the diagnosis was not confirmed. A weak relationship between plasma atrial natriuretic peptide (ANP) and left ventricular fractional shortening was demonstrated (r = -0.5, P < 0.001). Using an upper limit of ANP in the healthy elderly subjects of 62 pmol.ml-1 (mean + SD), plasma concentrations of ANP in the population with suspected heart failure had a sensitivity of 74% and specificity of 66% for the diagnosis of heart failure among elderly patients in the community or where access to echocardiography is limited. Left ventricular diastolic filling (assessed by Doppler) was abnormal in healthy elderly subjects and patients with heart failure, and appeared of limited value in the diagnosis of heart failure secondary to diastolic dysfunction. This study confirms that the renin-angiotensin system is activated in elderly patients with heart failure treated with diuretics. ANP may be helpful in diagnosing heart failure where it appears to have a complimentary role to echocardiography.