Sixteen very premature babies, born prior to term at gestational ages from 26–30 weeks, who required ventilation, were randomly assigned to groups, which did or did not receive treatment with surfactant. They were then studied prospectively with respect to the hemodynamic significance of the patency of the arterial duct. The velocity of end-diastolic flow in the celiac trunk was measured by Doppler echocardiography. Reduced values were assumed to represent the effect ofa large duct. The volume of systemic and pulmonary flows were estimated by means of Doppler echocardiographic studies at the tricuspid and mitral valvar orifices. Patency of the duct was diagnosed in all patients. Time to subsequent closure and the incidence of necrotizing enterocolitis and intracranial hemorrhage did not differ between the groups. The size of the shunt across the duct was measured as approximately 19% in patients with and without treatment. Velocity of end-diastolic flow, however, and the volumes of systemic flow, were markedly decreased in both, indicating the presence of an arterial duct ofhemodynamic significance. In contrast, in a control group of 20 normal newborns, end-diastolic flow velocities and systemic flows were normal when an arterial duct was diagnosed as patent. These results suggest that treatment with surfactant does not alter shunting across the arterial duct, but that gestational age may be the main factor influencing the hemodynamic significance of the duct. Due to the serious reduction of volumes of systemic flow in preterni babies with a patent duct, early closure should be considered in this population.