Myasthenia gravis (MG) is an autoimmune neuromuscular disorder due to autoantibodies directed against the neuromuscular junction. Conventional treatments are limited by hefty adverse effects mandating the set-up of novel therapies. Mesenchymal Stromal Cells (MSC) are multipotent progenitors that modulate immune and inflammatory responses through soluble factors, extracellular vesicles and cell-cell contacts. Our team demonstrated that the transfer of research grade (RG-) MSC conditioned by a co-culture step with peripheral blood mononucleated cells (PBMC) improved the clinical outcomes in a humanized MG mouse model (NSG-MG) (Sudres et al., JCI Insight, 2017). To shift towards a clinical perspective, RG-MSC were replaced by clinical grade (CG-) MSC (grown using human platelet lysate). Here, we aim to characterize the phenotypic and gene expression profiles of resting and conditioned CG-MSC, compared to a classical conditioning approach (IFN-γ activation), as well as to validate their functional capacities in vitro and in vivo. Therefore, CG-MSC derived from adipose tissue were conditioned by PBMC (cMSC), activated by IFN-γ (γ-MSC), or left untreated (rMSC). Their phenotype was assessed by flow cytometry (65 antibodies) and mass cytometry (CyTOF, Fluidigm) (31 antibodies). Gene expression was analyzed by RNA-seq (Illumina). Results showed that rMSC and cMSC had close phenotypic profiles whereas major phenotypic changes were found in γ-MSC. Some markers were modulated by all conditioning treatments, while others exclusively by one of them, thus conferring specific signatures. Gene expression profiles also differed between tested conditions, suggesting that different pathways are involved. In vitro immunomodulating capacities of CG-MSC supernatants were evaluated by PBMC proliferation inhibition assay and we observed higher responses upon cellular conditioning. Proteomic analyses (PEA methodology, Olink) were carried out using PBMC-MSC coculture supernatants and potential conditioning molecules were identified. Finally, the in vivo efficacy of CG cMSC was challenged in our humanized preclinical mouse model. cMSC-treated mice presented lower MG scores than untreated mice as early as 2 weeks post-injection, and a better survival. Our work unveiled treatment-dependent phenotypic markers of MSC, it demonstrates that immunomodulation capacities in vitro and in vivo are enhanced by cellular conditioning and it enlights their potential mechanisms of action