Clinical heart failure, often the result of myocardial infarction, may be preceded by a period of compensated left ventricular impairment. There is substantial need for an experimental model that reflects this human condition. In sheep, coronary artery ligation produced consistent left ventricular anteroapical myocardial infarctions resulting in chronic (5 wk), stable hemodynamic changes compared with sham controls, including reductions in ejection fraction (51 +/- 2 vs. 30 +/- 5%, P < 0.001), cardiac output (6.3 +/- 0.2 vs. 5.1 +/- 0.2 l/min, P < 0.01), and arterial pressure (93 +/- 2 vs. 79 +/- 3 mmHg, P < 0.001), and increases in cardiac preload (left atrial pressure, 3.3 +/- 0.1 vs. 8.3 +/- 1.3 mmHg, P < 0.001). These changes were associated with acute and sustained increases in plasma concentrations of atrial natriuretic peptide (ANP; 5 wk, 11 +/- 2 vs. 27 +/- 5 pmol/l, P < 0.001), brain natriuretic peptide (BNP; 3 +/- 0.2 vs. 11 +/- 2 pmol/l, P < 0.001), and amino-terminal pro-brain natriuretic peptide (NT-BNP; 17 +/- 3 vs. 42 +/- 12 pmol/l, P < 0.001). Significant correlations were observed between plasma levels of the natriuretic peptides (ANP, day 7 to week 5 samples; BNP and NT-BNP, day 1 to week 5 samples) and changes in left ventricular volumes and ejection fraction. In contrast, renin activity, aldosterone, catecholamines, and endothelin were not chronically elevated postinfarction and were not related to indexes of ventricular function. Coronary artery ligation in sheep produces the pathological, hemodynamic, and neurohormonal characteristics of compensated left ventricular impairment secondary to myocardial infarction. Plasma concentrations of the cardiac natriuretic peptides are sensitive markers of left ventricular dysfunction. This is a reproducible model that reflects the clinical condition and should prove suitable for investigating the pathophysiology of, and experimental therapies in, early left ventricular dysfunction.