BACKGROUND: The role of chemotherapy in addition to combined endocrine therapy for premenopausal women with endocrine-responsive early breast cancer remains an open question, yet trials designed to answer it have repeatedly failed to adequately accrue. The International Breast Cancer Study Group initiated two concurrent trials in this population: in Premenopausal Endocrine Responsive Chemotherapy (PERCHE), chemotherapy use is determined by randomization and in Tamoxifen and Exemestane Trial (TEXT) by physician choice. PERCHE closed with inadequate accrual; TEXT accrued rapidly. METHODS: From 2003 to 2006, 1317 patients (890 with baseline data) were randomly assigned to receive ovarian function suppression (OFS) plus tamoxifen or OFS plus exemestane for 5 years in TEXT. We explore patient-related factors according to whether or not chemotherapy was given using descriptive statistics and classification and regression trees. RESULTS: Adjuvant chemotherapy was chosen for 64% of patients. Lymph node status was the predominant determinant of chemotherapy use (88% of node positive treated versus 46% of node negative). Geography, patient age, tumor size and grade were also determinants, but degree of receptor positivity and human epidermal growth factor receptor 2 status were not. CONCLUSIONS: The perceived estimation of increased risk of relapse is the primary determinant for using chemotherapy despite uncertainties regarding the degree of benefit it offers when added to combined endocrine therapy in this population.