Objective Pre-treatment anxiety (PA) before chemotherapy increases complaints of chemotherapy-related symptoms (CRS). The re-sults on the association have been inconsistent, and the effect of temperament remains unclear. We aimed to determine whether PA is a risk factor for CRS and the effect of differing temperaments on CRS. Methods This prospective study comprised 176 breast cancer patients awaiting adjuvant chemotherapy post-surgery. We assessed CRS, PA, and temperament using the MD Anderson Symptom Inventory (MDASI), the Hospital Anxiety and Depression Scale, and the short form of the Temperament and Character Inventory-Revised, respectively. The MDASI was re-administered three weeks after the first chemo-cycle. Results PA showed weak positive correlation with several CRS after the first cycle; no CRS was significantly associated with PA when pre-treatment depressive symptoms and baseline CRS were adjusted in multiple regression analysis. Moderation model analysis indicated that the PA effect on several CRS, including pain, insomnia, anorexia, dry mouth, and vomiting, was moderated by harm avoidance (HA) but not by other temperament dimensions. In particular, PA was positively associated with CRS in patients with low HA. Conclusion The results in patients with low HA indicate that more attention to PA in patients with confident and optimistic tempera-ments is necessary.