Usefulness of the Doppler method under continuous infusion of adenosine triphosphate disodium (ATP) for improvement of accuracy in the diagnosis of the left internal thoracic arterial graft (LITA) patency was examined using transthoracic ultrasonic echocardiography. 1) Influence of ATP on the Doppler velocity in a graft was examined in 7 patients with good LITA grafts using physiological saline as the control. In the ATP group, 80 mg of ATP was dissolved in 20 ml physiological saline and continuously infused at 0.14 mg/kg/min. In the saline group, an equal volume of physiological saline was administered and the blood flow velocity in the LITA was recorded continuously by the transthoracic Doppler method from the supraclavicular fossa approach. Results; ATP administration increased the blood flow velocity in the LITA and the rate of increase was 48.3% for systolic peak velocity, 111% for diastolic peak velocity, 64.4% for systolic time velocity integral and 99% for diastolic time velocity integral indicating particularly high rates of increase in diastolic components. The diastolic/systolic peak velocity ratio or diastolic fraction did not increase significantly. In the saline group, none of the parameters showed a change. 2) Angiographic findings of the LITA were compared with the measurement values of the diastolic components by the Doppler method to examine usefulness of diastolic component measurement with ATP infusion for diagnosis of LITA patency. Subjects were 19 patients with good LITA (group A) and 8 patients with bad LITA (group B). Results; while there were significant differences in the mean baseline diastolic peak velocity, mean diastolic time velocity integral and mean diastolic fraction between the groups, overlapping was seen in individual cases. However, the inter-group differences were more distinct by ATP infusion and the borderline values were 30 cm/sec for diastolic peak velocity and 10 for diastolic time velocity integral. 3) Reliability of the diagnosis for LITA patency by measuring the diastolic components using the Doppler method with ATP infusion was examined and compared with the angiographic findings as the gold standard. Subjects were 27 patients and the diagnostic criteria for good LITA were set at 30 cm/sec for diastolic peak velocity and 10 for diastolic time velocity integral. Results; sensitivity and specificity of the Doppler method with ATP infusion were 100% for diagnosis of LITA patency by measuring the diastolic components. Conclusion, in diagnosis of LITA patency by the transthoracic ultrasonic cardiography, diagnostic accuracy was improved by measuring the diastolic parameters under continuous infusion of ATP.