The purpose of the study was to assess the prognosis of HER2-positive metastatic breast cancer patients who received trastuzumab beyond progression and investigate the predictors of complete response. HER2-positive metastatic breast cancer patients who received long-term trastuzumab were included in the study. Predictors of complete response were analyzed with binary regression analysis. The prognosis of patients who had their trastuzumab-based treatment terminated was assessed. Eighty patients were involved in the study. The patients were received with trastuzumab for a median of 62 months (12-191). A complete response was observed in 60 (75%) patients. The median duration to development of complete response was found as 14.8 months (2.4-55). In logistic regression analysis: using endocrine therapy with trastuzumab (p = 0.04), menopausal status (p = 0.03), and the number of metastatic sites (p = 0.01) were found to be statistically significant factors for a complete response. Trastuzumab-based therapy of fifteen patients was terminated, six (40%) patients continued to receive an aromatase inhibitor, and nine (60%) patients were followed up without treatment. After termination of trastuzumab, at a median follow-up of 32 months (11-66), recurrence was detected in two (13.3%) patients. We detected that menopausal status, the number of metastatic sites, and using endocrine therapy with trastuzumab were predictors of complete response in HER2-positive metastatic breast cancer patients who received long-term trastuzumab-based therapy. We observed that HER2-positive metastatic breast cancer patients may be completely cured with trastuzumab-based therapy. There are no defined criteria for termination of trastuzumab treatment in this selected patient group. It is necessary to confirm our data with multicenter studies involving a large number of patients.