BACKGROUND Limited data exist on optimal empiric oral antibiotic treatment for outpatients with cellulitis in areas with a high prevalence of community-associated methicillin-resistant Staphylococcus aureus (MRSA) infections. METHODS We conducted a 3-year retrospective cohort study of outpatients with cellulitis empirically treated at a teaching clinic of a tertiary-care medical center in Hawaii. Patients who received more than 1 oral antibiotic, were hospitalized, or had no follow-up information were excluded. Treatment success rates for empiric therapy were compared among commonly prescribed antibiotics in our clinic: cephalexin, trimethoprim-sulfamethoxazole, and clindamycin. Risk factors for treatment failure were evaluated using multivariate logistic regression analysis. RESULTS Of 544 patients with cellulitis, 405 met the inclusion criteria. The overall treatment success rate of trimethoprim-sulfamethoxazole was significantly higher than the rate of cephalexin (91% vs 74%; P<.001), whereas clindamycin success rates were higher than those of cephalexin in patients who had subsequently culture-confirmed MRSA infections (P=.01), had moderately severe cellulitis (P=.03), and were obese (P=.04). Methicillin-resistant S. aureus was recovered in 72 of 117 positive culture specimens (62%). Compliance and adverse drug reaction rates were not significantly different among patients who received these 3 antibiotics. Factors associated with treatment failure included therapy with an antibiotic that was not active against community-associated MRSA (adjusted odds ratio 4.22; 95% confidence interval, 2.25-7.92; P<.001) and severity of cellulitis (adjusted odds ratio 3.74; 95% confidence interval, 2.06-6.79; P<.001). CONCLUSIONS Antibiotics with activity against community-associated MRSA, such as trimethoprim-sulfamethoxazole and clindamycin, are preferred empiric therapy for outpatients with cellulitis in the community-associated MRSA-prevalent setting.