Part 1 of our report, presented in the same issue of the Journal, shows that immediate postoperative extubation and direct transfer to the surgical ward is safe and reduces reliance on the intensive care unit in most liver transplant recipients. However, there is no method to preoperatively predict which patients will need ventilatory support after surgery. To address this issue, we examined the relationship between perioperative patient attributes and extubation outcome in patients entered into our immediate postoperative extubation study from 1996 to 1998. Variables chosen stemmed from considerations in the literature. We examined the influence of 13 preoperative and 6 intraoperative factors on extubation outcome. Preoperative attributes included sex, race, diagnosis, United Network for Organ Sharing status, Child-Pugh score, presence of a portosystemic shunt, ascites, encephalopathy, coagulation, age, body mass index (BMI), creatinine level, and year of surgery. Intraoperative factors were type of surgery, surgeon, anesthesiologist, number of red blood cells administered, length of surgery, and surgical start time. Female sex (P =.02), BMI of 32 or greater (P =.015), portosystemic shunt (P =.022), and encephalopathy (P =.041) were associated with no attempt by the physician to extubate, whereas encephalopathy (P =.01) and BMI of 34 or greater (P =.002) were associated with failure to meet criteria for postoperative extubation (described in part 1 of this study). We conclude there are limited factors that predict an increased risk for postoperative respiratory failure in liver transplant recipients. Our results indicate that physicians are conservative in their approach to extubation immediately after surgery, and sole reliance on physician judgment to determine suitability for postoperative extubation leads to unnecessary use of postoperative cardiopulmonary support.