Mesenchymal stromal cells (MSCs) have been extensively utilised in clinical trials to treat a wide range of diseases. However, their mechanism of action remains poorly understood, as there is little evidence that these cells survive in vivo administration. Therefore we sought to understand how MSCs mediate therapeutic effects that persist beyond their survival. We showed that MSCs localised to the lungs immediately after intravenous administration, where they underwent caspase 3-dependent apoptosis within 1 hour. Apoptotic MSCs were efferocytosed by lung phagocytes, which then shut down inflammation. Immunosuppression did not require MSCs to remain viable, as chemically induced apoptotic MSCs elicited immunosuppressive effects to a similar extent as viable MSCs. Treatment of MSCs with pharmacological compounds that selectively inhibit regulators of apoptosis revealed that human and mouse MSCs utilised different pro-survival proteins. Having identified the molecular players involved in MSC apoptosis and survival, we generated apoptosis-resistant MSCs via CRISPR/Cas9 technology. Utilising disease models commonly used to define MSC potency, we were able to assay the therapeutic efficacy of these apoptosis-resistant MSCs to answer the outstanding question of how short-lived MSCs are immunosuppressive. Our data have significant implications for the development of MSC potency assays and biomanufacturing for clinical translation.