At rest, during cardiac catheterization, aortic plasma levels of immunoreactive atrial natriuretic peptide did not differ between 10 controls with atypical chest pains and normal coronary arteries and 9 patients with stable angina pectoris and coronary arterial disease (55.2 +/- 19.8 vs. 64.8 +/- 19.8 pg/ml, NS). Nor did atrial natriuretic peptide values differ between the two groups during or after atrial pacing (150 beats/minute), which induced electrocardiographic and metabolic signs of acute myocardial ischaemia in the patients with coronary arterial disease but in none of the controls. Pacing, when carried out for more than 300 seconds, induced an increase of plasma atrial natriuretic peptide that correlated with duration of pacing (r = 0.80, P less than 0.001), and similarly in controls and patients with coronary arterial disease. In a second part of the study, which included 2 controls and 2 patients with coronary arterial disease, post-pacing coronary sinus concentrations of atrial natriuretic peptide were 10-20 times higher than peripheral levels (415- greater than 890 pg/ml). The concentration of atrial natriuretic peptide rose as blood from the caval veins (34 +/- 7 pg/ml) entered the right atrium (56 +/- 24 pg/ml), but thereafter was unchanged in the pulmonary artery (51 +/- 3 pg/ml) and the aorta (46 +/- 9 pg/ml). In conclusion, the results gave no evidence for ischaemic heart disease without congestive cardiac failure to be associated with altered levels of atrial natriuretic peptide. It was confirmed that atrial pacing stimulates the secretion of atrial natriuretic peptide which is produced by the heart and released via the coronary sinus into the circulation.