Background: The FN rate for the approved regimen of TC is 5% in pivotal studies. Other small retrospective reports have reported FN rates as high as 20-35%. We report the incidence of FN from a large retrospective series of breast cancer patients receiving TC with or without pegfilgrastim (PF) for adjuvant therapy. Methods: We reviewed records of 240 sequential patients who had received adjuvant TC (75 and 600 mg/m²) between Mar ’07 and Nov ’12 for FN, upfront PF use, and adverse events by treatment cohort. FN was defined as T> 100.4°F and ANC <500, while upfront PF was defined as PF given at physician discretion 24-48h after 1^stcycle of TC administration. Comparisons between two proportions used exact binomial methods; effect of PF on FN after adjusting for baseline characteristics was tested using multivariate logistic regression. Results: 153 (63.7%) patients received upfront PF, while 87 (36.3%) did not (Table). Patients receiving upfront PF were older (57 vs. 52 yrs, p=0.02). Other baseline characteristics (size of primary tumor, hormone receptor status, and nodal status) were no different between groups. FN, fever, antibiotic use, hospitalization, and dose delay were significantly higher when upfront PF was not used (Table). The average length of hospital stay was 2.9 days for TC + PF, and 3.8 days for TC pts. 31/87 (35.6%) of patients without upfront PF went on to receive PF after the 1st cycle. Conclusions: FN for adjuvant TC meets clinical practice threshold and ASCO guidelines for upfront use of PF. FN-related outcomes such as fever, antibiotic use, dose delays, and number of hospitalizations, and were significantly increased without upfront PF. The cost-effectiveness of these findings will be presented and have major clinical implications for routine care.