BACKGROUND: We sought to determine the proportion of community-associated Staphylococcus aureus infections due to methicillin-resistant S. aureus (CA-MRSA) at a large county hospital. In addition, we sought to identify the demographic and clinical risk factors associated with CA-MRSA infection. METHODS: Patients were prospectively enrolled if they were admitted to Parkland Hospital and had a positive culture for S. aureus isolated within 72 h of admission. The patients were interviewed using a standardized data questionnaire. Data collected included patient demographics, clinical history, as well as health care and non-health care associated MRSA risk factors. Bacterial susceptibilities were verified through review of microbiology laboratory and pharmacy records. Isolates were tested for Panton-Valentine leukocidin (PVL) gene, SCCmec type, and for inducible clindamycin resistance. RESULTS: One hundred and ninety-eight patients were interviewed prospectively, of which eight had colonization without active infection. One hundred and nineteen patients were infected with MRSA and 71 patients were infected with methicillin-susceptible S. aureus (MSSA). Patients with MRSA were more likely to be African-American and unemployed. Patients with MRSA most commonly presented with a skin or soft tissue infection (SSTI): 69% versus 45%, p=0.0012, while patients with MSSA were more likely to have infection of the respiratory tract: 11% versus 3%, p=0.02. Patients with MRSA were more likely to have used antibiotics in the past six months, been homeless, have a history of incarceration, have abused alcohol and have a history of infection with MRSA. In multivariate analysis, African-American race, antibiotics in the past six months, and a history of being homeless were associated with MRSA infection. Only 11 of 119 (9%) MRSA patients did not have at least one of these risk factors. PVL gene was present in 72 of 74 (97%) MRSA isolates and SCCmec type IV was present in 63 of 75 (84%) MRSA isolates. CONCLUSIONS: The majority of patients hospitalized with community-associated S. aureus infections were due to MRSA, most of which involved an SSTI. African-American race, recent antibiotics and past homeless status predicted infection with MRSA; however, no clinical profile could reliably exclude MRSA. Clinicians should be aware of the increasing prevalence of CA-MRSA.