Objective To evaluate whether using long axis (LA) or short axis (SA) view during ultrasound-guided internal jugular (IJ) and subclavian (SC) central venous catheterization (CVC) results in fewer skin breaks, decreased time to cannulation, and fewer posterior wall penetrations (PWP). Design Prospective, randomized crossover study. Setting Urban emergency department with approximate annual census of 60,000. Subjects Emergency medicine resident physicians at the Denver Health Residency in Emergency Medicine, a PGY 1-4 training program. Interventions Resident physicians blinded to the study hypothesis used ultrasound guidance to cannulate the IJ and SC of a human torso mannequin using the LA and SA views at each site. Measurements An ultrasound fellow recorded skin breaks, redirections, and time to cannulation. An experienced ultrasound fellow or attending used a convex 8–4 MHz transducer during cannulation to monitor the needle path and determine PWP. Generalized linear mixed models with a random subject effect were used to compare time to cannulation, number of skin breaks and redirections, and PWP of the LA and SA at each cannulation site. Results 28 resident physicians participated: 8 PGY-1, 8 PGY-2, 5 PGY-3, and 7 PGY-4. The median [interquartile range (IQR)] number of total IJ central venous catheters placed was 27 (IQR 9-42) and SC was 6 (IQR 2-20) catheters. The median number of previous ultrasound-guided IJ catheters was 25 (IQR 9-40), and ultrasound-guided SC catheters was 3 (IQR 0-5). The LA view was associated with a significant decrease in the number of redirections at the IJ and SC sites, relative risk (RR) 0.4 (95% confidence interval [CI] 0.2-0.9), and RR 0.5 (95% CI 0.3-0.7), respectively. There was no significant difference in the number of skin breaks between the LA and SA at the SC and IJ sites. The LA view for SC was associated with decreased time to cannulation; there was no significant difference in time between the SA and LA views at the IJ site. The prevalence of PWP was: IJ SA 25%, IJ LA 21%, SC SA 64%, and SC LA 39%. The odds of PWP were significantly less in the SC LA, odds ratio 0.3 (95% CI 0.1-0.9). Conclusions The LA view for the IJ was more efficient than the SA view with fewer redirections. The LA view for SC CVC was also more efficient with decreased time to cannulation and fewer redirections. The LA approach to SC CVC is also associated with fewer PWP. Using the LA view for SC CVC and avoiding PWP may result in fewer central venous catheter-related complications.