Breast cancer is the most common cancer amongst women worldwide, however clinically validated chemotherapy response biomarkers that can accurately predict treatment response in patients are largely lacking. Therefore, in this study, functional paclitaxel and eribulin ex vivo sensitivity assays were developed. Patient derived xenograft (PDX) models were used to compare the ex vivo predicted treatment outcome with the sensitivity of mice in vivo. We validated the previously developed sensitivity assay for paclitaxel, which is based on the ratio between replicating (EdU) and mitotic (phospho-Histone H3; pH3) cells as a proxy for blocked mitosis. The assay showed 90% correlation between the ex vivo and in vivo response to paclitaxel treatment in the PDX models. We propose the term REMIT (REplication MITosis) assay and show that it is also a suitable test to predict eribulin sensitivity. The reproducibility of the REMIT assay for paclitaxel and eribulin was determined to be 80% and 83%, respectively. These results justify further clinical validation of the REMIT assay in breast cancer patients.