In order to assess the value of exercise thallium scintigraphy for the detection and prognosis of graft coronary artery disease, 50 heart transplant patients (mean age 46.7 +/- 11.5 years) were studied within 48 h of their scheduled yearly coronary angiography and subsequently followed up for a mean of 13 +/- 3 months. Angiography revealed normal coronary arteries in 35 patients, and coronary artery disease in 15 (two with type A lesions, seven with type B lesions and six with both). Seven patients had one or more stenoses > or = 50%. Exercise thallium scintigraphy was negative in all patients with normal coronary arteries (100% specificity), and abnormal in 10 of 15 patients with coronary artery disease (67% sensitivity). Fixed defects were seen in six cases, transient defects in two and both in two; the results of the test were abnormal in all seven patients with > or = 50% lesions. During follow-up, none of the patients with a normal exercise thallium scintigraphy experienced any cardiac event; in the group with abnormal results, four cardiac events occurred. Although further studies are needed to confirm these results, exercise thallium scintigraphy seems to be useful in evaluating post-transplant coronary artery disease: it is accurate in detecting the most severe stenoses and provides some prognostic information.