Many patients show physiological signs of worsening for several hours prior to requiring more aggressive interventions and transfer to a higher level of care. Rapid response teams have been widely deployed to address this problem, but this approach is fundamentally reactive rather than proactive and has had mixed results so far. This time series study utilized an electronic physiological surveillance system—a real-time decision support system based on patients' vital signs—embedded within the electronic medical record to provide guidance for clinicians in determining patients at risk for deterioration and optimizing treatment intensity. Implementation of the electronic physiological surveillance system was associated with a statistically significant reduction in mortality for a broad range of diagnoses at both hospitals. The results of this study illustrate the potential of novel information technology approaches for prospectively identifying patients at risk for clinical harm.