Background: Studies of risk factors for clinically significant pressure ulcers in the hospital have been limited by the small number of study subjects that develop pressure ulcers, resulting in contradictory findings regarding some risk factors. Objective: To determine if three risk factors (low serum albumin level, fecal incontinence, and confusion) were significant risk factors when tested in a large data set. Methods: The study design was a longitudinal cohort study using data collected as a component of a multi-site controlled clinical trial. The data were collected at 47 Veterans Affairs Hospitals. 2,771 subjects that required high levels of nursing care were identified to have mobility impairment. Their medical records were abstracted using a standard form to identify a large number of potential risk factors. The subsequent development of stage 2 or greater pressure ulcers was recorded for a maximum of 14 days after admission. Results: 406 patients (14.7%) subsequently developed at least one stage 2 or greater pressure ulcer over a 2-week period. In a multivariate model, the presence of low albumin levels (odds ratio OR = 1.40) and confusion (OR = 1.45) were both found to be statistically significant risk factors, while fecal incontinence was not. Having a Do Not Resuscitate (DNR) order was also a significant risk factor (OR = 1.55). Two other known risk factors also entered the model: being malnourished (OR = 1.69) and requiring a urinary catheter (OR = 1.55). Conclusions: This study confirmed confusion and low albumin as pressure ulcer risk factors, but not fecal incontinence. A DNR order was found to be a new pressure ulcer risk factor not previously described in the literature.