Cancer patients undergo detrimental toxicities and ineffective treatments especially in the relapsed setting, due to failed treatment attempts. The development of a tool that predicts the clinical response of individual patients to therapy is greatly desired. We have developed a novel patient-derived 3D tissue engineered bone marrow (3DTEBM) technology that closely recapitulate the pathophysiological conditions in the bone marrow and allows ex vivo proliferation of tumor cells of hematologic malignancies. In this study, we used the 3DTEBM to predict the clinical response of individual multiple myeloma (MM) patients to different therapeutic regimens. We found that while no correlation was observed between in vitro efficacy in classic 2D culture systems of drugs used for MM with their clinical efficacious concentration, the efficacious concentration in the 3DTEBM were directly correlated. Furthermore, the 3DTEBM model retrospectively predicted the clinical response to different treatment regimens in 89% of the MM patient cohort. These results demonstrated that the 3DTEBM is a feasible platform which can predict MM clinical responses with high accuracy and within a clinically actionable time frame. Utilization of this technology to predict drug efficacy and the likelihood of treatment failure could significantly improve patient care and treatment in many ways, particularly in the relapsed and refractory setting. Future studies are needed to validate the 3DTEBM model as a tool for predicting clinical efficacy.