A high incidence of coronary artery fistulas has been observed angiographically after heart transplantation. To determine the present incidence of this finding and the natural history of fistulas in this setting, we reviewed coronary angiograms and clinical course on all patients (n = 480) transplanted from 1980 to 1990 who survived until the first annual coronary angiogram and compared the incidence of coronary artery fistulas in the early (patients #1-160), middle (patients #161-320), and late transplant (patients #321-480) groups. The 3-yr coronary artery fistula incidence for the early group was 3.5%, 6.9% for the middle group, and 2.9% in the late group (P < 0.05, early vs late). Patients who developed fistulas were followed longitudinally. Angiographic follow-up data (median duration: 6 yr) were available in 14 patients having 17 fistulas. No fistula increased in size, and in 10 of 14 patients (71%), fistulas became angiographically undetectable. No patient had any clinical complication related to the fistula. In summary, the incidence of coronary artery fistulas is presently lower than previously reported, which may in turn be related to refinements in heart biopsy technique. The lack of long-term clinical sequelae and the relatively high rate of disappearance favor a conservative approach of "watchful waiting."