Animal reports suggest that reflex activation of cardiac sympathetic nerves can evoke coronary vasoconstriction. Conversely, physiologic stress may induce coronary vasodilation to meet an increased metabolic demand. Whether the sympathetic nervous system can modulate coronary vasomotor tone in response to stress in humans is unclear. Coronary blood velocity (CBV), an index of coronary blood flow, can be measured in humans by noninvasive duplex ultrasound. We studied 11 healthy volunteers and measured beat-by-beat changes in CBV, blood pressure, and heart rate during: a) static handgrip for 20 s at 10% and 70% of maximal voluntary contraction; b) lower body negative pressure at -10 and -30 mmHg for 3 min each; c) cold pressor test for 90 s; d) hypoxia (10% O₂), hyperoxia (100% O₂) and hypercapnia (5% CO₂) for 5 min each. At the higher level of handgrip, mean blood pressure increased (P < 0.001) while CBV did not change (P = NS). In addition, during lower body negative pressure CBV decreased (P < 0.02; and P < 0.01, for -10 and -30 mmHg, respectively) while blood pressure did not change (P = NS). The dissociation between the responses of CBV and blood pressure to handgrip and lower body negative pressure is consistent with coronary vasoconstriction. During hypoxia CBV increased (P < 0.02) and decreased during hyperoxia (P < 0.01) although blood pressure did not change (P = NS), suggesting coronary vasodilation during hypoxia and vasoconstriction during hyperoxia. In contrast, concordant increases in CBV and blood pressure were noted during the cold pressor test, and hypercapnia had no effects on either parameter. Thus, physiologic stress known to be associated with sympathetic activation can produce coronary vasoconstriction in humans. Contrasting responses were noted during systemic hypoxia and hyperoxia where mechanisms independent of autonomic influences appear to dominate the vascular end organ effects.