Background: Despite treatment with high-dose chemotherapy followed by autologous stem cell transplantation (AHCT), MM patients invariably relapse. MRD-negativity post-AHCT has emerged as the most important prognostic marker. Currently, MRD in MM is monitored via bone marrow aspirate sampling. Marrow MRD assays are limited by the spatial heterogeneity of marrow MM localization; extramedullary disease and sampling variability of marrow aspiration. Sensitive, non-invasive blood-based MRD assay is an unmet need. ctDNA as a noninvasive biomarker can be utilized to predict relapse in MM. Here we attempt to evaluate MRD using ctDNA in AHCT recipients with MM. Methods: In this retrospective, single-center study, we analyzed ctDNA MRD in blood samples collected from 28 patients with MM after upfront AHCT. A total of 80 plasma timepoints were available pre and post AHCT with a median follow-up of 92.4 months. Multiparameter flow cytometry (MFC) at 10-4 level was used to assess the MRD from the BM biopsy. Individual bone marrow aspirates or FFPE slides from the time of MM diagnosis and matched normal blood were whole-exome sequenced, and somatic mutations were identified. MRD assessment at 3 months post-AHCT was performed by ctDNA analysis using a personalized, tumor-informed (Signatera^TM, bespoke mPCR NGS assay). The prognostic value of ctDNA was evaluated by correlating MRD status with clinical outcomes. Results: Table provides the baseline disease characteristics. Median age was 67 [41-75] years and 16 [57.1%] were males. ctDNA was detectable in 70.8% (17/24) of pre-AHCT, 53.6% (15/28) of ̃3 months post-AHCT, and 39.2% (11/28) of patients during the surveillance phase post-AHCT. Of the 15 ctDNA MRD positive patients, 93.3% (n=14) experienced relapse on follow-up (hazard ratio: 5.64; 95% CI: 1.8-17; p=0.0003). Patients negative for ctDNA at 3 months post-AHCT had significantly superior progression-free survival (PFS) compared to positive (median PFS, 84 months vs. 31 months; p=0.003) The positive predictive value (PPV) for relapse among patients positive for ctDNA at 3 months post-AHCT was 93.3%, and significantly higher than marrow MFC of 68.4%. Conclusions: Our study shows the feasibility that a tumor-informed assay on archival blood samples is predictive of relapse post-AHCT. Future prospective studies with real-time marrow NGS and ctDNA samples are needed to define the role of ctDNA in MM and its prognostic significance.