BACKGROUND: Internal jugular vein access is an essential part of patient management in many clinical settings and is usually achieved with a blind, anatomical landmark-guided technique. The purpose of this study is to evaluate whether a 2-dimensional ultrasound technique can improve on the traditional method. METHODS: Eighty patients undergoing elective open heart surgery were randomly assigned to an anatomical landmark group or ultrasound group (each n = 40). With the patient in the supine position, the head was rotated 30o to the contralateral side and triangle formed by the clavicle and both heads of the sternocleidomastoid muscle were identified. We accessed the internal jugular vein from the apex of the triangle toward the ipsilateral nipple in the anatomical landmark group. The internal jugular vein and carotid artery were visualized with two-dimensional ultrasound. We compared the number of advances made with the central venous cannulation needle, the time to blood aspiration, complications and failure rate. RESULTS: The failure rate was 22.5% using the anatomical landmark technique and 5% using the ultrasound technique. The vein was entered on the first attempt in 60% of patients using the landmark technique and in 63% using ultrasound (P>0.05). Mean attempts for puncture were 1.9 (anatomical) vs 1.6 (ultrasound-guided) (P>0.05). Complications occurred in 20% of cases using anatomical landmarks and in 5% using ultrasound (P>0.05). The average access time was 26.5 seconds by the anatomical approach and 56 seconds by the ultrasound approach (P< 0.05). CONCLUSIONS: Ultrasound-guided cannulation of the internal jugular vein did not significantly decrease failure rate, access time, complications, and attempts for puncture, nor did it increase the amount of successful first attempts. However an investigation using a larger number of patients will be needed.