Computerized provider order entry (CPOE) has been enthusiastically endorsed as a means of preventing medication errors, and some studies have shown clinical benefits associated with its use. However, a growing body of literature demonstrates that CPOE implementation may have many unintended consequences affecting clinician workflow, and these issues may account for the mixed safety performance of CPOE. This study, conducted at an academic tertiary care hospital, found that overall inpatient mortality decreased after CPOE was launched, with an estimated 36 lives saved in the first 18 months after implementation. These findings counter an earlier, widely cited study in which mortality increased after CPOE installation, and point to the importance of careful attention to the implementation process to ensure that CPOE meets its potential.