Atrial filling fraction, or the fraction of stroke volume resulting from atrial contraction, was measured by Doppler echocardiography from the time-velocity integral of mitral anulus inflow with a method that allows separation of conduit or passive flow from flow resulting from the atrial contraction. The method was validated in 17 patients with externally programmable ventricular demand pacemakers by showing that the time-velocity integral of passive flow (excluding the A wave) during sinus or sequential atrioventricular pacing was almost identical to the time-velocity integral during ventricular pacing. Atrial filling fractions were then measured in 41 normal subjects, aged 20 to 80 years; 28 patients with echocardiographic evidence of concentric left ventricular hypertrophy; 24 with dilated cardiomyopathy (13 of whom had an ischemic origin); and 19 with acute myocardial infarction. Atrial filling fraction increased significantly with age in normal subjects (r = 0.77; p less than 0.001) and ranged from 12% in a 20-year-old man to 46% in a normal 80-year-old woman. In the hypertrophy group, atrial filling fraction had a weak relation with age (r = 0.47; p = 0.006), and the values were significantly higher than in normal subjects. In patients with cardiomyopathy or infarction, atrial filling fraction varied over a wide range and showed no relation to age. Thus, atrial filling fraction as determined by Doppler echocardiography is significantly altered by both age and left ventricular disease. Age-corrected nomograms are essential when assessing atrial filling fraction in individual patients.