Chemotherapy is vital for breast cancer management, but early onset toxicities like neutropenia hinder its administration. Primary prophylactic (PP) use of colony- stimulating factors (CSF) helps prevent neutropenia and ensures successful chemotherapy completion. Nevertheless, lack of specific guidelines for CSF administration in older patients has lead to unexplained geographic and racial, and counter-intuitive clinical variations in CSF administration. This study examined the reasons for these variations and for the first time looked at variations in PP-CSF administration and duration of administration in breast cancer patients receiving chemotherapy. This retrospective observational study of newly diagnosed breast cancer patients receiving chemotherapy used SEER-Medicare data from 1994–2003. Regression analyses were used to explore the factors associated with PP-CSF administration and duration of administration. Clinical and therapeutic characteristics previously unexplored by other studies were included. Univariate analyses demonstrated geographic, racial and clinical disparities similar to previous studies. However, clinical correlations resolved to statistical insignificance after inclusion of chemotherapy characteristics. The analysis showed that significant geographic and racial disparities existed. History of recent antibiotic use was associated with shorter PP-CSF administration. Physicians’ decision to administer PP-CSF is predominantly driven by neutropenia risk associated with pre-planned chemotherapy regimen. Older, sicker women at a higher risk of neutropenia receive less intense/toxic chemotherapy and thus do not require PP-CSF. Geographic variations are driven by proportion of physicians administering PP-CSF with no evidence for overuse among specific physicians. Association of recent antibiotic use with shorter PP-CSF administration suggests intended substitution of the expensive PP-CSF with prophylactic antibiotics.