To compare the quality of lung collapse, time, and number of attempts required to achieve lung isolation, and incidence of intraoperative malpositioning between the EZ blocker (EZB), Fuji Uniblocker (UB), and the left-sided double lumen tube (DLT). Prospective, randomized clinical trial. Single tertiary-level, university-affiliated hospital. Eighty-nine patients undergoing elective open thoracotomies or video-assisted thoracoscopic surgery. The 89 patients were randomized to receive a DLT, UB, or EZB for one-lung ventilation. The quality of lung collapse at the time of pleural opening and 10 and 20 minutes thereafter were assessed by the surgeon using the Lung Collapse Score (LCS; 0=no lung collapse to 10=best lung collapse). The time and number of attempts required to achieve lung isolation and the number of repositions required during surgery were measured. Tracheobronchial tree measurements were performed by radiologists from preoperative computed tomography imaging. The surgeon remained blinded to the type of device used. Twenty-nine patients were randomized to the DLT group and 30 patients to each of the EZB and UB groups. The LCSs among the groups at pleural opening and 10 minutes after pleural opening were not significantly different (p=0.34 and p=0.08, respectively). However, at 20 minutes after the pleural opening, the LCSs were significantly different among groups (p=0.02), with median scores being significantly lower for DLT (9 [IQR 8-9]) than for EZB (9 [IQR 9-10]; p=0.04) and UB (9.5 [IQR 9-10]; p=0.02). Lung isolation was achieved fastest in the DLT group (p < 0.01). The frequency of difficult placement did not significantly differ among groups, although it occurred most frequently in UB (n=7; 23.3%). Intraoperative repositioning also occurred most often with the UB (n=15; 50.0%). The EZB had the greatest number of cases requiring >2 repositions (n=4, 13.3%). There were no differences between preoperative airway measurements and time to isolation or incidence of intraoperative repositioning among the groups. The LCS was comparable among the 3 devices until 20 minutes after pleural opening, when better scores were obtained in the bronchial blocker groups. Lung isolation was achieved fastest with the DLT. The EZB had the highest incidence of cases requiring >2 intraoperative repositions, mostly occurring in R-sided surgery. For L-sided surgery, the EZB performed equally to the UB. This suggests that using the EZB for R-sided video-assisted thoracoscopic surgery may be suboptimal. Preoperative airway dimensions did not correlate with time to achieve isolation or incidence of intraoperative malpositioning.