Objective: We wish to present the preliminary result of transarterial occlusion of patent ductus arteriosus (PDA) with Gianturco coils in pediatric patients in central Taiwan. Materials and Methods: We attempted occlusion of PDA with Gianturco coils in a total of 26 consecutive patients, 13 infants and 13 children, 23 female and three male, between July 1 1997 to September 30 1998. Median patient age was 2.57 years (from 0.25 to 14.02 years old). Median patient weight was 10.8 kg (4.0 to 36.0 kg). Premature babies with PDA, full-term babies who were less than three months old and patients who had other congenital heart disease were not included in this study. All PDAs were approached transarterially from the femoral artery. Coils were selected to provide a helical diameter that was twice or more the minimum ductus diameter and a length approximating five loops. In five patients who had a PDA diameter ≥3.5 mm, we used a snare technique to assist coil delivery beforehand, and to test coil stability, or to retrieve coil that had migrated to the pulmonary artery afterwards. Physical auscultation, chest radiographs and echocardiography with color Doppler were done in all patients within 24 h, and one, two, three, six and 12 months after coil occlusion. Results. The median ductus minimum diameter was 2.3 mm (range, 1.0 to 4.7 mm). Fifteen patients had the megaphone type (type A), four had the window type (type B), five had the tubular type (type C), one had the aneurysmal type (type D) and one had the elongated conical type (type E). Twenty-one patients underwent single coil occlusion and five had multiple coils occlusion. Twenty-one patients had immediate angiographic closure of the ductus and disappearance of heart murmur at 15 min after the procedure. Dark-brown urine (hemoglobinuria) was found in one patient, 10 h after the first procedure, due to a mild residual ductal shunt. Two more coils were implanted in a second procedure that was performed within 24 h, and the ductus was completely occluded. The dark-brown urine regressed. At one month follow-up, four patients had mild residual ductal shunts, which were completely occluded by one more coil in three patients and by two more coils in the other patient. Malpositioned coils were deployed in five patients immediately after the procedure. In total, the closure rate at 15 min, within 24 h, and at one, two, three, six and 12 months were 81, 85, 85, 100, 100, 100 and 100%, respectively. In one year of follow-up, there was no instance of coil migration, ductus reopening or stenosis of the left pulmonary artery. Conclusions. Transarterial occlusion of PDA, with a Gianturco coil having approximately five loops, can be effectively achieved in patients with a minimum ductus diameter up to 4.7 mm. In patients with a ductus of more than 3.5 mm, the snare-assisted technique was employed advantageously to control coil delivery with accuracy and stability. Coil malposition or migration can be easily retrieved using a 10-mm Nitnol snare catheter. Hemoglobinuria, due to intravascular hemolysis, may regress within 24 h after the second attempt at coil implantation.