Objective To explore the effect of PICC localization by ultrasound and traditional chest X-ray in preterm infants. Methods A self-control study was conducted on 116 preterm infants. PICC localization was confirmed by ultrasound and radiograph, and indexes were observed and recorded immediately. Results The detection rate of tip position was 99.14% by ultrasound and 100% by X-ray. The rates of PICC tip location at the entrance of the superior and inferior caval veins into the right atrium were 100%, 100% by ultrasound and 98.28%, 95.69% by X-ray. The success rate of measuring the distance between catheter tip and the entrance of the atrium was 99.14% by ultrasound and 93.97% by X-ray. Consistency of the two evaluation techniques was high, with Kappa=0.806 and Spearman correlation coefficient = 0.753 (P＜0.01 for both). Interquartile range of ultrasonic measurement data was 0.482 cm, and chest radiograph measurement data was 0.642 cm. The number of X-ray exposure was 145. The operation time of ultrasonic localization was significantly shorter than that of chest radiography localization （P＜0.01）. Conclusion Ultrasound, a safe and easy to operate technique, directly shows the position of catheter tip, so it is an accurate and stable method for premature infants.