BACKGROUND: The effectiveness of surveillance to identify extended-spectrum beta-lactamase-producing Enterobacteriaceae (ESBL-E) carriers is controversial during a non-outbreak situation. We performed additional stool cultures for ESBL-E among intensive care unit (ICU) patients already under active surveillance by means of sputum and urine cultures. We aimed to assess the efficacy of stool cultures for screening for ESBL-E in a non-outbreak situation. METHODS: We conducted a retrospective cohort study in an ICU. Sputum and urine samples were cultured for ESBL-E surveillance purposes from January to September 2013 (phase 1). Stool cultures were routinely performed in addition from January to September 2014 (phase 2). Antimicrobial use density values and clinical outcomes were investigated and compared between phase 1 and 2. RESULTS: We identified 512 and 478 patients in phase 1 and phase 2, respectively. ESBL-E were found in the feces of 65 (13.6%) patients in phase 2. The antimicrobial use density values (expressed as defined daily doses per 1,000 bed-days) were not significantly different between the two phases for fluoroquinolones (7 vs. 10, p = 0.376), third-generation cephalosporins (24.2 vs. 29.5, p = 0.724), tazobactam/piperacillin (44.6 vs. 57.3, p = 0.489), and carbapenems (73 vs. 55.5, p = 0.222). Moreover, there were no significant differences in ICU mortality and length of stay (11.5% vs. 9.8%, p = 0.412, and 9 vs. 10 days, p = 0.28, respectively). CONCLUSIONS: Stool culture seemed ineffective in improving the antimicrobial use density of broad-spectrum antimicrobials, clinical outcomes, and ICU length of stay, and is not recommended for surveillance of ESBL-E in a non-outbreak situation.