Background Pregnancy alters the anatomic landmarks for internal jugular vein (IJV) cannulation. In this study, IJV cannulation was simulated, and success of the technique was evaluated using ultrasound. Methods Term pregnant women and non-pregnant female volunteers were recruited. The degree of difficulty in assessing neck anatomy was noted. The optimal insertion points for cannulation were marked on the skin of each subject according to the central landmark and palpatory techniques. The midpoint of a 15-10 MHz linear transducer was placed over each insertion point, and the vertical cursor of the ultrasound, which represented the path of the needle, was placed in the image. The outcome was an IJV puncture, a carotid artery (CA) puncture, or a missed attempt. Results One hundred and sixty-one women, 99 pregnant and 62 non-pregnant, were studied. The identification of landmarks was more difficult in pregnant women (P = 0.01). The rates of successful IJV punctures, CA punctures, and missed attempts did not differ significantly between pregnant and non-pregnant women. Carotid artery punctures using the central landmark technique in pregnant and non-pregnant subjects were 19% and 10%, respectively. Corresponding figures for the palpatory approach were 6% and 3%, respectively. The degree of IJV overlap of the CA in the palpatory technique was greater in pregnant women (38% vs 18% showed [ 75% overlap; P \ 0.001), a difference that remained after controlling for body mass index Conclusion The IJV overlies the CA to a greater extent in pregnant patients than in non-pregnant patients. Thus, a landmark approach for IJV cannulation might expose pregnant women to a greater risk of carotid puncture. This trial was registered at www.clinicaltrials.gov (NCT T00464828).