Human epidermal growth factor receptor 2 (HER2) is an effective therapeutic target in breast cancer. However, not all patients benefit from trastuzumab-based therapy. We aimed to investigate whether patients with different levels of HER2 amplification would experience different clinical outcomes with trastuzumab-based chemotherapy. We quantified the HER2 gene copy number (GCN) and HER2/centromere chromosome probe 17 (CEP17) ratio in 291 breast cancer patients with HER2 amplification confirmed by immunohistochemistry and fluorescence in situ hybridization. The optimal cutoff points for HER2 GCN and the HER2/CEP17 ratios for distinguishing positive results were determined by receiver operating characteristic (ROC) curve analyses. ROC analysis identified optimal cutoff points for HER2 GCN and HER2/CEP17 ratios as 11.5 and 6.5 (P = 0.039 and P = 0.012), respectively. The DFS in patients with HER2 GCN <11.5 was significantly longer than in HER2 GCN ≥11.5 patients (P = 0.015) according to Kaplan-Meier survival curves analysis. Similarly, patients with HER2/CEP17 ratios <6.5 had a significantly longer DFS than those with HER2/CEP17 ratios ≥6.5 (P = 0.013). Moreover, patients with HER2 cluster amplification showed a worse survival than those with HER2 non-cluster amplification (P = 0.041). This study demonstrated a significant association between the level of HER2 amplification and survival time in a relatively large cohort of HER2-positive breast cancer patients undergoing trastuzumab-based chemotherapy. Further investigations of more precise quantitative measurements and larger cohorts are required to define this threshold.