The purpose of this study was to determine if a decolonization regimen reduces the frequency of methicillin-resistant Staphylococcus aureus (MRSA) infections and if colonization isolates are genetically related to subsequent infectious strains. Trauma patients admitted to the intensive care unit with positive MRSA nasal swabs were randomized to either daily chlorhexidine gluconate (CHG) baths and mupirocin (MUP) ointment to the nares or soap and water baths and placebo ointment for five days. Nasal swabs performed at the end of treatment and invasive MRSA infections during the remaining hospitalization were compared with the original nasal isolate via polymerase chain reaction for genetic relatedness as well as CHG and MUP resistance genes. Six hundred and seventy-eight intensive care unit admissions were screened, and 92 (13.6%) had positive (+) MRSA nasal swabs over a 22-month period ending in 3/2014. After the five day treatment period, there were 13 (59.1%) +MRSA second nasal swabs for CHG + MUP and 9 (90%) for soap and water baths and placebo, P = 0.114. No isolates tested positive for the MUP or CHG resistance genes mupA and qacA/B but 7 of 20 (35%) contained smr. There were seven (31.8%) MRSA infections in the CHG group and six (60%) for soap, P = 0.244. All 13 patients with MRSA infections had the same MRSA isolate present in the original nasal swab. There was no difference in all-cause Gram-negative or positive infections for CHG versus soap, 12 (54.5%) versus 7 (70%), P = 0.467. CHG + MUP are ineffective in eradicating MRSA from the anterior nares but may reduce the incidence of infection. Subsequent invasive MRSA infections are typically caused by the endogenous colonization strain.