There is still controversy as to the manner in which Doppler-derived transmitral filling patterns change because of myocardial ischemia. To evaluate the effects of exercise-induced ischemia on Doppler-derived filling patterns, 28 patients were examined at rest and during three stages of supine bicycle exercise (0.5, 1.0, and 1.5 W/kg). The peak early (E) and integrated early (Ei) and peak late (A) and integrated late (Ai) diastolic flow velocities, as well as their ratios (E/A, Ei/Ai), were compared between patients with exercise-induced ischemia but no wall-motion abnormalities at rest (ischemia group, n = 13) and those with akinetic scars from previous infarction but no exercise ischemia (scar group, n = 15). Normal subjects with no evidence of heart disease served as a control group (n = 11). At maximal workload the ischemia group showed a significantly lower peak flow velocity at atrial contraction than the control and scar group (0.74 +/- 0.18 vs 1.08 +/- 0.25 and 0.89 +/- 0.19 m/sec, respectively; p < 0.05) and also a significantly lower flow velocity integral at atrial contraction (8.24 +/- 2.2 vs 12.81 +/- 4.8 and 11.32 +/- 3.6 cm, respectively; p < 0.05). Therefore, the atrial contribution to filling was diminished during ischemia (36.2% +/- 9.2% vs 47.3% +/- 6.4% and 48.4% +/- 13.8%, respectively; p < 0.05). By maintaining the early filling rate during ischemia, the reduced atrial contribution resulted in a significantly higher E/A ratio (1.48 +/- 0.31 vs 1.05 +/- 0.15 and 1.16 +/- 0.44, respectively) and Ei/Ai ratio (2.0 +/- 1.06 vs 1.09 +/- 0.26 and 1.24 +/- 0.79, respectively). The assessment of Doppler-derived transmitral filling during exercise-induced ischemia shows mainly early diastolic filling, which is in contrast to the profile of impaired relaxation usually associated with ischemia. Evidence of exercise-induced ischemia leading to greater increases in left atrial pressure suggests that transmitral filling patterns are more closely related to hemodynamic status than to diastolic function.