The National Institute for Health and Care Excellence (NICE) invited the manufacturer of eribulin (Eisai Ltd) to submit evidence for the clinical and cost effectiveness of eribulin as treatment for patients with locally advanced or metastatic breast cancer (LABC/MBC) pre-treated with at least two chemotherapy regimens. This article summarizes the review of evidence by the Evidence Review Group (ERG) and provides a summary of the NICE Appraisal Committee's (AC's) decision. The clinical evidence was derived from a multi-centred, open-label, randomized, phase III study comparing eribulin with treatment of physician's choice (TPC) in 762 patients with LABC/MBC. Clinical effectiveness results were submitted for two populations: the overall intention-to-treat (ITT) population and a subset (n = 488) that included only patients from North America, Western Europe and Australia (Region 1). For the primary endpoint of overall survival (OS), a primary analysis (after 55 % of patients had died) and an updated analysis (after 77 % of patients had died) were conducted. In the ITT population, treatment with eribulin was associated with a significant improvement in median OS compared with TPC in both primary [difference in median OS 2.5 months; hazard ratio (HR) 0.81, 95 % confidence interval (CI) 0.66-0.99] and updated analyses (2.7 months; HR 0.81, 95 % CI 0.67-0.96). A statistically significant improvement in progression-free survival (PFS) was reported for eribulin compared with TPC when assessed by the investigator (difference in median PFS 1.48 months; HR 0.76, 95 % CI 0.64-0.90), but not when assessed by the ERG (1.44 months; HR 0.87, 95 % CI 0.71-1.05). Gains in OS were greater for Region 1 patients than for the ITT population (3.1 vs. 2.7 months). Health-related quality of life (HRQoL) data suggested a benefit for eribulin responders, but was based on phase II studies. In the eribulin arm, serious adverse events included febrile neutropenia (4.2 %) and neutropenia (1.8 %), with peripheral neuropathy being the most common reason for treatment discontinuation. The manufacturer's economic evaluation using Patient Access Scheme costs reported a base-case incremental cost-effectiveness ratio (ICER) for eribulin versus TPC (Region 1) of £46,050 per quality-adjusted life year gained (corrected to £45,106 when an erroneous data entry was removed). The ERG's revised ICERs were £61,804 for Region 1 and £76,110 for the overall population. The AC concluded that the evidence had not demonstrated sufficient benefit in OS, cost effectiveness or HRQoL and that eribulin was not recommended for use in this patient group.