A system to provide surveillance, diagnosis, and protocolized management of surgical intensive care unit (SICU) sepsis was undertaken as a performance improvement project. A system for sepsis management was implemented for SICU patients using paper followed by a computerized system. The hypothesis was that the computerized system would be associated with improved process and outcomes. A system was designed to provide early recognition and guide patient-specific management of sepsis including (1) modified early warning signs-sepsis recognition score (MEWS-SRS; summative point score of ranges of vital signs, mental status, white blood cell count; after every 4 hours) by bedside nurse; (2) suspected site assessment (vascular access, lung, abdomen, urinary tract, soft tissue, other) at bedside by physician or extender; (3) sepsis management protocol (replicable, point-of-care decisions) at bedside by nurse, physician, and extender. The system was implemented first using paper and then a computerized system. Sepsis severity was defined using standard criteria. In January to May 2012, a paper system was used to manage 77 consecutive sepsis encounters (3.9 ± 0.5 cases per week) in 65 patients (77% male; age, 53 ± 2 years). In June to December 2012, a computerized system was used to manage 132 consecutive sepsis encounters (4.4 ± 0.4 cases per week) in 119 patients (63% male; age, 58 ± 2 years). MEWS-SRS elicited 683 site assessments, and 201 had sepsis diagnosis and protocol management. The predominant site of infection was abdomen (paper, 58%; computer, 53%). Recognition of early sepsis tended to occur more using the computerized system (paper, 23%; computer, 35%). Hospital mortality rate for surgical ICU sepsis (paper, 20%; computer, 14%) was less with the computerized system. A computerized sepsis management system improves care process and outcome. Early sepsis is recognized and managed with greater frequency compared with severe sepsis or septic shock. The system has a beneficial effect as a clinical standard of care for SICU patients. Therapeutic study, level III.