Background. Nasal colonization with methicillin-resistant Staphylococcus aureus (MRSA) can be a precursor to serious infection, and decolonization with topical mupirocin has been studied as a means of preventing clinical infection. Mupirocin resistance in patients with MRSA has been reported, usually in the context of widespread mupirocin use. Methods. Patients admitted to a surgical intensive care unit (SICU) had nasal swab cultures for MRSA performed at admission, weekly, and at discharge in an active surveillance program. Collected MRSA isolates were tested for mupirocin resistance, and molecular analysis was performed. Clinical data on the characteristics and outcomes of the patients who stayed in the SICU for >48 h were collected prospectively. Results. Of the 302 MRSA isolates available for testing, 13.2% were resistant to mupirocin, with 8.6% having high-level resistance (minimum inhibitory concentration, ≥512 μg/mL) and 4.6% having low-level resistance (minimum inhibitory concentration, 8-256 μg/mL). Patients admitted to the SICU for >48 h who were colonized with mupirocin-resistant MRSA were more likely to have been admitted to our hospital during the previous year (P = .016), were older (P = .009), and had higher in-hospital mortality (16% vs. 33%; P = .027), compared with patients colonized with mupirocin-susceptible MRSA. Molecular analysis of the mupirocin-resistant isolates revealed that 72.5% of isolates contained staphylococcal cassette chromosome mec II. Repetitive sequence polymerase chain reaction typing revealed that high-level mupirocin resistance was present in multiple clonal groups. The rate of mupirocin use hospital-wide during the study period was 6.08 treatment-days per 1000 patient-days. Conclusions. We documented a high rate of mupirocin resistance in MRSA isolates from SICU patients, despite low levels of in-hospital mupirocin use.