Background: Despite many years of investigation into mesenchymal stem cells (MSCs) and their potential for treating inflammatory conditions such as COVID-19, clinical outcomes remain variable due to factors like donor variability, different tissue sources, and diversity within MSC populations. Variations in MSCs’ secretory and proliferation profiles, and their proteomic and transcriptional characteristics significantly influence their therapeutic potency, highlighting the need for enhanced characterization methods to better predict their efficacy. This study aimed to evaluate the biological characteristics of MSCs from different tissue origins, selecting the most promising line for further validation in a K18-hACE2 mouse model of SARS-CoV-2 infection. Methods: We studied nine MSC lines sourced from either bone marrow (hBMMSC), dental pulp (hDPMSC), or umbilical cord tissue (hUCMSC). The cells were assessed for their proliferative capacity, immunophenotype, trilineage differentiation, proteomic profile, and in vitro immunomodulatory potential by co-culture with activated lymphocytes. The most promising MSC line was selected for further experimental validation using the K18-hACE2 mouse model of SARS-CoV-2 infection. Results: The analyzed cells met the minimum criteria for defining MSCs, including the expression of surface molecules and differentiation capacity, showing genetic stability and proliferative potential. Proteomic analysis revealed distinct protein profiles that correlate with the tissue origin of MSCs. The immunomodulatory response exhibited variability, lacking a discernible pattern associated with their origin. In co-culture assays with lymphocytes activated with anti-CD3/CD28 beads, all MSC lines demonstrated the ability to inhibit TNF-α, to induce TGF-β and Indoleamine 2,3-dioxygenase (IDO), with varying degrees of inhibition observed for IFN-γ and IL-6, or induction of IL-10 expression. A module of proteins was found to statistically correlate with the potency of IL-6 modulation, leading to the selection of one of the hUCMSCs as the most promising line. Administration of hUCMSC to SARS-CoV-2-infected K18 mice expressing hACE2 was effective in improving lung histology and modulating of a panel of cytokines. Conclusions: Our study assessed MSCs derived from various tissues, uncovering significant variability in their characteristics and immunomodulatory capacities. Particularly, hUCMSCs demonstrated potential in mitigating lung pathology in a SARS-CoV-2 infection model, suggesting their promising therapeutic efficacy.