Background: A phase II, multicenter, randomized clinical trial of EG and PG as first-line chemotherapy for patients with HER2-negative MBC showed that EG was less neurotoxic, but conferred a survival outcome similar to that of PG. In this study, we analyzed FACT-T questionnaires from patients participating in this clinical trial to determine their health-related Quality of life(HR QoL). Methods: Patients were randomly assigned to either EG or PG chemotherapy arm in a 1:1 ratio. QoL was assessed using the Korean version of the FACT-T questionnaire. After baseline assessment, HRQoL was assessed every 2 cycles for 12 cycles and every 3 cycles after 13 cycles of chemotherapy. The linear mixed model was used to evaluate the difference in HRQoL between the two treatments. Results: Of 118 patients, 117 patients except 1 patient in PG arm responded to the FACT-T questionnaires at baseline. Baseline FACT-T subscale QoL scores and overall QoL scores were not different between the EG and PG arms. During treatment, overall QoL scores and other FACT-T subscale scores did not differ between EG and PG arm. In terms of taxane-associated HRQoL, PG arm much increased taxane subscale scores after 2 cycles of chemotherapy compared to EG arm until the 13^th cycle of treatment ( all ps < 0.05, except 11^th cycle [p = 0.164]). Of taxane subscale scores, neuropathy specific subset scores were presented as similar pattern to taxane subscale scores. After 13 cycles of treatment, both groups had similarly intense symptoms. Therefore, although taxane subscale score and neuropathy specific subset score were higher in the PG arm compared to the EG arm, there was no statistical significance (p = 0.086 and p = 0.062, respectively). Conclusions: EG delayed and decreased chemotherapy induced adverse events including neuropathy compared to PG. Therefore, eribulin would be a reasonable substitute for paclitaxel as first line treatment in MBC, especially concerning neurotoxicity. Clinical trial information: NCT02263495