Background: The 18F-fluoroestradiol positron emission tomography/computed tomography (18F-FES PET/CT) technique provides a convenient method to evaluate the overall estrogen receptor (ER) expression in metastatic breast cancer (MBC) patients. There are long debates on the characteristics and treatment strategy of patients with positive primary ER lesion but negative ER expression in metastatic disease. 18F-FES PET offers an opportunity to answer this question. Methods: This study enrolled MBC patients with ER-positive primary tumor who received 18F-FES PET/CT in our center. Descriptive statistics were used in clinicopathologic characteristics and compared with Chi square test or t test. Progression‐free survival (PFS) was estimated by Kaplan-Meier method and compared by log-rank test. Results: 16.46% (52/314) patients with an ER-positive primary tumor had negative ER expression assessed by 18F-FES for MBC prior to receiving first-line systemic therapy. The rate of ER-negativity tested by 18F-FES was negatively co-related with levels of ER expression in early stage, with a positive to negative conversion in 2 of 3 (46.3%) ER-low (ie, ER 1%-9%) tumors, 7 of 10 (70.0%) ER-moderate (ie, ER 10%-49%) tumors, 16 of 114 (14.0%) ER-high (ie, ER 50%-95%) tumors, and 1 of 8 (12.5%) ER-very high (ie, ER > 95%) tumors (p < 0.001). Chemotherapy (83.33%, 40/48) was the most common treatment strategy afterward, among which capecitabine monotherapy (64.29%, 27/42) was a dominant alternative. PFS was significantly prolonged with capecitabine alone versus other chemotherapy (median PFS: 13.14 vs 6.21 months, p = 0.035). Conclusions: 18F-FES could identify negative conversion of ER in MBC which occurred frequently. Patients with lower ER expression in the primary lesion were more likely to have negative ER expression in the metastasis. In real-world clinical practice, chemotherapy was the primary choice by most physicians and capecitabine monotherapy had shown good efficacy.