Background The epidemiology of methicillin-resistant Staphylococcus aureus (MRSA) maybe changed by strict infection control measures, and the impact of empirical antibiotic therapy on the outcomes of MRSA infection was not clear. We aimed to investigate the present epidemiological status of MRSA infection and empirical antibiotic therapy for MRSA infection in university teaching hospitals in mainland China. Methods The present study was a multicenter prospective observational study conducted in five university teaching hospitals. Patients who were consecutively admitted to the intensive care unit and signed a consent form from March 3, 2011 to May 31, 2011 were included. Patients with age <18 years or with a length of hospital stay <48 hours were excluded from this study. The following variables were collected or recorded: demographic data, general status, APACHE II score of the patient at the time of admission, infections, and the use of antibiotics during a stay. Primary outcomes and prognostic indicators included length of hospital stay and 28-day and 90-day mortality. The differences between the patients with appropriate empirical therapy and patients with inappropriate therapy were analyzed to detect the influences of antibiotic therapy on the prognosis of MRSA infection. Results A total of 682 cases were enrolled. Thirty (66.2%) of 88 MRSA cases were treated with effective antibiotics for MRSA infection; only 20% received appropriate empirical antibiotic treatment. The empirical therapy group compared with the target therapy group had a shorter length of stay, but there were no significant differences in mortality rates. There were no significant differences in the length of hospital stay, length of stay, and 28-day and 90-day mortality between MRSA-infected patients who received or not received effective antibiotics. Two hundred and eighteen cases received sensitive antibiotics for MRSA. Conclusions The MRSA infection rates are at relatively low levels in university teaching hospitals in China. The empirical use of sensitive antibiotics for MRSA infection was at relatively high rate, and there is a tendency of overusing in patients without MRSA infection. On the other hand, the rate of appropriate empirical antibiotic therapy for patients with MRSA infection is relatively low.