Background Mesenchymal stem cells (MSCs), also known as mesenchymal stromal cells, have been proposed as a promising therapeutic option for people with multiple sclerosis on the basis of their immunomodulatory and neuroprotective properties. The MEsenchymal StEm cells for Multiple Sclerosis (MESEMS) study was devised to evaluate the safety, tolerability, and activity of autologous MSCs derived from bone marrow and infused intravenously in patients with active multiple sclerosis. Methods MESEMS is a randomised phase 2 trial done at 15 sites in nine countries. Patients (aged 18-50 years) with active relapsing-remitting or progressive multiple sclerosis were included if they had a disease duration of 2-15 years since onset of multiple sclerosis and an Expanded Disability Status Scale score of 2.5-6.5. Patients were randomly assigned (1:1), according to a crossover design, to receive a single intravenous dose of autologous bone marrow derived MSCs followed by placebo at week 24, or to receive placebo followed by autologous MSCs at week 24, with a follow-up visit at week 48. Primary objectives were to test safety and activity of MSC treatment. The primary safety endpoint was to assess the number and severity of adverse events within each treatment arm. The primary efficacy endpoint was the number of gadolinium-enhancing lesions (GELs) counted over week 4, 12, and 24 compared between treatment groups. The primary efficacy endpoint was assessed in the full analyis set, after all participants completed the week 24 visit. Efficacy endpoints were evaluated using a predefined statistical testing procedure. Safety was monitored throughout the study by recording vital signs and adverse events at each visit. Findings From July 16, 2012, until July 31, 2019, 144 patients were randomly assigned to first receive early intravenous infusion of autologous bone marrow-derived MSCs (n=69) or placebo (n=75). MSC treatment did not meet the primary endpoint of efficacy on the total number of GELs accumulated from baseline to week 24 (rate ratio [RR] 0.94, 95% CI 0.58-1.50; p=0.78). 213 adverse events were recorded, similarly distributed between groups (93 cases recorded in 35 [51%] of 69 patients treated first with MSCs vs 120 cases in 42 [56%] of 75 patients infused first with placebo). The most frequent adverse events reported were infection and infestations, with a total of 54 (25%) of 213 adverse events (18 [19%] of 93 in the early-MSC group and 36 [30%] of 120 in the delayed-MSC group). Nine serious adverse events were reported in seven patients treated with placebo versus none in the MSC group. All serious adverse events were considered to be unrelated to the treatment infusion. No deaths were reported during the study. Interpretation Bone marrow-derived MSC treatment was safe and well tolerated but did not show an effect on GELs, an MRI surrogate marker of acute inflammation, in patients with active forms of multiple sclerosis, at week 24. Thus, this study does not support the use of bone marrow-derived MSCs to treat active multiple sclerosis. Further studies should address the effect of MSCs on parameters related to tissue repair. Copyright (C) 2021 Elsevier Ltd. All rights reserved.