The mechanism or mechanisms that cause intrapulmonary arteriovenous anastomoses (IPAVA) to either open during exercise in subjects breathing room air and at rest when breathing hypoxic gas mixtures, or to close during exercise while breathing 100% oxygen, remain unknown. During conditions when IPAVA are open, plasma epinephrine (EPI) and dopamine (DA) concentrations both increase, potentially representing a common mechanism. The purpose of this study was to determine whether EPI or DA infusions open IPAVA in resting subjects breathing room air and, subsequently, 100% oxygen. We hypothesized that these catecholamine infusions would open IPAVA. We performed saline-contrast echocardiography in nine subjects without a patent foramen ovale before and during serial EPI and DA infusions while breathing room air and then while breathing 100% oxygen. Bubble scores (0-5) were assigned based on the number and spatial distribution of bubbles in the left ventricle. Pulmonary artery systolic pressure (PASP) was estimated using Doppler ultrasound, while cardiac output (Q(C)) was measured using echocardiography. Bubble scores were significantly greater during EPI infusions of 80-320 ng·kg(-1)·min(-1) compared with baseline when subjects breathed room air; however, bubble scores did not increase when they breathed 100% oxygen. At comparable Q(C) and PASP, intravenous DA (16 μg·kg(-1)·min(-1)) and EPI (40 ng·kg(-1)·min(-1)) resulted in identical bubble scores. Subsequent studies revealed that β-blockade did not prevent hypoxia-induced opening of IPAVA. We suggest that increases in Q(C) or PASP (or both) secondary to EPI or DA infusions open IPAVA in normoxia. The closing mechanism associated with breathing 100% oxygen is independent from the opening mechanisms.