Background. Both dipyridamole and adenosine are widely used as pharmacologic stressors with Tl-201 imaging for detection of coronary artery disease. The put-pose of this study was to compare dipyridamole and adenosine Tl-201 imaging directly in patients with angiographically proved coronary artery disease. Methods and Results. Fifty-four patients were submitted to two planar Tl-201 studies: one with dipyridamole and the other with adenosine. The interval between the two studies varied from 2 to 7 days and the order was assigned randomly. Three standard planar views were obtained 10 minutes and 4 hours after the injection of 3.0 mCi Tl-201. Administration of dipyridamole was as follows: 0.142 mg/kg/min during 4 minutes, followed by a slight exercise and Tl-201 injection. The infusion of adenosine was as follows: 0.140 mg/kg/min during 6 minutes with injection of Tl-201 after the third minute of infusion. Patients were asked to give their preference considering the number type, severity, and duration of side effects on a scale from 0 (worst) to 5 (best). Reading was done by two experienced observers. The heart was divided into three segments per view. The change in systolic blood pressure was -12+/-11 mm Hg for adenosine and -5 +/- 10 mm Hg for dipyridamole (p < 0.001), and the change in heart rate was 18 +/- 10 beats/min for adenosine and 8 +/- 7 beats/min for dipyridamole (p < 0.001). With regions of interest, ischemic/normal wall ratios were determined: 0.78 +/- 0.06 for adenosine and 0.83 +/- 0.08 for dipyridamole (p < 0.001). Adenosine detected 295 normal, 170 ischemic, and 21 scar segments, whereas dipyridamole detected 326, 135, and 25 segments, respectively. Patients preferred adenosine (4.3 +/- 1.0 for adenosine vs 3.8 +/- 1.5 for dipyridamole; p < 0.04) mainly because of the short duration of side effects. Conclusion. This study shows that the use of adenosine with Tl-201 imaging may have some advantages over dipyridamole.