Background, Combining low-level treadmill exercise,vith adenosine infusion may result in fewer bradyarrhythmic complications by increasing sympathetic tone and may improve myocardial perfusion image quality by decreasing background activity. Methods. Patients referred for outpatient pharmacologic stress myocardial perfusion imaging performed simultaneous treadmill exercise (mean 2.2 metabolic equivalents) throughout 6-minute adenosine infusion (adenosine-exercise n = 507), Patients unable to exercise and those with left bundle branch block received adenosine infusion alone (adenosine-nonexercise n = 286), Adverse reaction data were collected on all patients and compared by sex. Background-to-target activity was calculated in a blinded fashion on 200 randomly selected patients. Results. During the period from April 1996 to December 1998, 507 patients (64%) underwent adenosine-exercise testing, whereas 286 (36%) underwent adenosine-nonexercise testing. Hypotensive and arrhythmic (atrioventricular block, sinus bradycardia, and new onset atrial fibrillation or flutter) adverse reactions occurred less often during adenosine-exercise than during adenosine-nonexercise. Neither death nor myocardial infarction occurred in either group. In the adenosine-exercise group, 2.8% of patients experienced an adverse reaction versus 5.6% of the adenosine-nonexercise group (P = .04), The reduction in adverse reactions occurred in both men and women, although women had significantly more adverse reactions than men (5.7% vs 1.8%, P = .004). Liver/heart and gut/heart ratios were lower in the adenosine-exercise group (1.05 +/- 0.42 vs 1.21 +/- 0.55, P = .01; 0.61 +/- 0.21 vs 0.69 +/- 0.24, P = .03, respectively). Conclusions. Compared with adenosine infusion alone, combining low-level treadmill exercise with adenosine in outpatients is safe, better tolerated, and improves image quality. Women were more likely to experience adverse reactions than men.