OBJECTIVES: Few studies have examined excess health-care costs among elderly breast cancer (BC) patients by receipt of human epidermal growth factor receptor 2- (HER2-) targeted therapy.  METHODS: Women aged 65+ with an incident diagnosis of BC (index) and no history of other cancer were identified from 2006-2010 linked Surveillance, Epidemiology, and End Results (SEER) and Medicare data. Women were divided into two cohorts based on receipt of HER2-targeted therapy (trastuzumab or lapatinib) and matched 1:1 to non-cancer comparison cohorts by age, sex, and race.  Continuous enrollment from 1 year pre-index (baseline) through disenrollment, death, or the end of the data was required. All-cause costs were evaluated per-patient-per-month (PPPM) overall and by stage.  Generalized linear models were constructed to identify factors associated with costs, by stage, controlling for demographics and comorbidity. RESULTS: We identified 1,746 BC patients receiving and 35,114 not receiving HER2-targeted therapy. Unadjusted excess total costs (vs. non-cancer patients) were $4,079 PPPM for the HER2-targeted cohort and $990 for the no HER2-targeted cohort (both P < 0.001), with larger differences at more advanced stages.  Excess cost drivers were outpatient care and physician/provider services (including HER2-targeted therapy acquisition and administration costs).  In multivariate analyses, Stage I HER2-targeted BC patients experienced 3.17 times greater total costs than non-cancer patients; while those with Stages II, III, and IV had 3.02, 3.40, and 4.27 times greater costs respectively (all P < 0.001). Similar trends with generally smaller magnitudes were observed among patients without HER2-targeted therapy (0.48 [I]; 0.83 [II]; 1.67 [III]; 4.33 [IV]; all P < 0.001). Other significant cost predictors included older age, Black or Hispanic race, and baseline Charlson score>2. CONCLUSIONS: Women with BC experience higher costs than non-cancer patients, with greater burden among those receiving HER2-targeted therapy.  Excess cost drivers were outpatient care and physician/provider services.