Abstract Objectives: This study aimed to evaluate the clinical utility of serum leukocyte counts (SLCs) for initial antibiotic use in Chinese women with mastitis who presented to the emergency department. Materials and Methods: Electronic medical records of breastfeeding women with mastitis were reviewed. Patients were divided into two groups according to the level of SLCs: slight elevated leukocyte counts (SELC, 1.0-1.5*10^9 cells/L); marked elevated leukocyte counts (MELC, >1.5*10^9 cells/L ). Treatment outcomes including rates of treatment failure and breast abscess formation were compared. Results: The rates of treatment failure and breast abscess were 12.7% and 7.1% in overall population and 7.7% and 6.4% in MELC group, respectively. In SELC group, treatment failures were observed in 29.7% and 6.4% patients without or with antibiotics, and the difference was significant (OR=4.207, 95% CI 1.318- 13.424); breast abscess was observed in 12.5% and 2.1% patients, and the difference was not significant (OR=6.571, 95% CI 0.793-54.481). Mean time to normal appearance of the breast or normal temperature was shorter in patients with antibiotics (3.8 ± 1.7 and 4.3 ± 2.1 days) than in patients without antibiotic use (2.5 ± 1.1 and 3.0 ± 1.3 days), p<0.001. Conclusion: Our study found that there were better clinical outcomes in patients with SELC or MELC when antibiotic was initially administrated compared to those without antibiotics, whicha indicates that SLC is an easy and practical reference index for gudiing antibiotic use, and patients who have an elevated SLC should be treated with antibiotic. Key Words: mastitis, antibiotic, serum leukocyte count