BACKGROUND: Eribulin was approved for patients with metastatic breast cancer (MBC) following anthracycline and taxane therapy based on modest survival gains in trial populations. We considered survival outcomes and 30-day post-chemotherapy mortality among all NHS patients under Cancer Drugs Fund eligibility criteria, treated with eribulin in England between January 2012 and August 2021. METHODS: Overall survival (OS) was estimated from start of eribulin using Poisson and Kaplan-Meier methods. Associations between patient characteristics and mortality were considered using Poisson and binomial regression. Thirty-day mortality was defined from the start of the most recent chemotherapy cycle. RESULTS: 6842 patients with MBC who initiated eribulin were identified. 88% of patients died during follow-up. Median OS was 8.9 months (95% CI 8.7-9.2); Survival was longer in patients with ER-positive disease (9.7 vs 7.4 months - compared to ER-negative) and strongly associated with performance status (PS0: 11.0; PS2+: 5.3 months). Adjusted mortality was lower in older patients, and substantially higher in those with poorer performance status (PS2+ vs PS0 RR 1.96, 95% CI 1.77-2.16). Overall, 17.2% patients died within 30 days of a cycle of chemotherapy, with higher risks in patients with poorer performance status, of younger age, and shorter interval since primary diagnosis. CONCLUSIONS: In routine NHS care in England, survival following eribulin was shorter and early mortality higher than reported in clinical trials, particularly among patients with impaired performance status. Careful patient selection and transparent discussion of risks are essential.