Background: This study aimed to examine the effectiveness and safety of eribulin used as an early-line (EL, i.e., first-/second-line) versus late-line (LL, i.e., third-line and beyond) chemotherapy for recurrent advanced or metastatic breast cancer (A/MBC) patients. Methods: This study conducted a retrospective observation of A/MBC patients initiating eribulin between January 1, 2015, and June 30, 2019, using medical database at a university-affiliated teaching hospital in Taiwan. Patients were assigned into either the EL or LL group based on the timing of respective eribulin treatments and were observed for at least 6 months up to December 2019 for progression-free survival (PFS), time to treatment failure (TTF), overall survival (OS), disease response, and occurrence of adverse events. The Kaplan-Meier and Cox proportional hazard regression survival analyses were performed. Results: Of 127 patients, 23.6% (n = 30) and 76.4% (n = 97) were assigned to the EL and LL groups, respectively, between which no difference in patient characteristics was noted. Median PFS and TTF were 6.5 months and 5.0 months for the EL and 4.2 months and 3.4 months for the LL, respectively. Median OS could not be estimated in the EL group and was 20.5 months in the LL group. Eribulin as an EL treatment was the only factor associated with longer TTF and OS, whereas the number of metastatic sites was additionally associated with PFS in the multivariate analysis. No complete response was reported in either group, but a partial response was obtained in 6.7% in the EL group and 3.1% in the LL group. The common adverse events between two groups were similar, including leukopenia (80.0%), neutropenia (76.7%), and anemia (60.0%). Conclusions: The eribulin used as an EL of chemotherapy was effective for A/MBC patients with known toxicities in this study, while eribulin as the LL chemotherapy showed consistent results with previous reports.