PURPOSE: Although the 2016 Japanese guidelines for the management of sepsis recommend de-escalation of treatment after identification of the causative pathogen, adherence to this practice remain unknown. The objective of this study was to evaluate the benefits of de-escalating treatment for sepsis patients at an advanced critical care and emergency medical centre. METHODS: Based on electronic patient information, 85 patients who were transported to the centre by ambulance, and diagnosed with sepsis between January 2008 and September 2013 were enrolled and evaluated. Patients were divided into two groups with and without de-escalation, and comparisons were conducted for several variables, including length of hospital stay, and length of antibiotic administration. Two types of subgroup analysis were conducted between patients with septic shock or positive blood cultures. Statistical analysis was conducted using chi-square and Mann-Whitney U tests. RESULTS: The length of hospital stay after diagnosis was significantly shorter for the de-escalation group than for the non-de-escalation group. In the subgroup analysis, de-escalation for blood culture-positive patients was beneficial in terms of the length of hospital stay and length of antibiotic administration. CONCLUSIONS: The findings of this study suggest that sepsis treatment de-escalation is beneficial for treatment efficacy and appropriate use of antibiotics. This article is open to POST-PUBLICATION REVIEW. Registered readers (see "For Readers") may comment by clicking on ABSTRACT on the issue's contents page.