Background Mesenchymal stromal cells (MSCs) have been shown in rodent models to promote primary and pulmonary metastatic sarcoma growth when injected in the presence of gross tumor. In theory, this would limit their use in a clinical setting after limb salvage treatment for osteosarcoma. Although concerning, these models do not translate to the clinical setting wherein MSCs could be used after primary tumor resection to aid in bone healing and incorporation of tumor endoprostheses. If we can determine whether the use of MSCs in this setting is safe, it might improve our ability to augment bone healing in patients undergoing limb salvage. Questions/purposes The purpose of this study was to determine (1) whether MSCs promote pulmonary metastatic disease progression in a murine osteosarcoma model; and/or (2) whether they affect local disease recurrence in the presence of microscopic residual osteosarcoma. Methods An orthotopic model of luciferase-expressing osteosarcoma was developed. At 10 days, resection of the primary tumor was performed. One hundred fourteen female C3H mice were inoculated with DLM8-luc osteosarcoma in the proximal tibia. Ninety-four mice developed orthotopic osteosarcoma with luciferase expression. Mice with bioluminescent evidence of a primary tumor received either a microscopically ''clean'' amputation at a time when residual microscopic metastatic disease was present in the lungs (pulmonary metastasis group; n = 65) or a ''dirty'' amputation (local recurrence group; n = 29). Mice were randomized to receive intravenous MSCs, MSCs at the surgical site, or no MSCs. Mice were monitored for development and progression of pulmonary metastasis and local recurrence by bioluminescence imaging and daily measurements at the surgical site. The number of pulmonary nodules, time to first evidence of metastasis, and size of recurrent tumor were compared using Kruskal-Wallis, analysis of variance, Welch's, t-tests, or Mann-Whitney tests as appropriate for the specific data sets with p \ 0.05 considered significant. Results Mice receiving intravenous MSCs had a faster time to first detection of pulmonary metastasis (2.93 ± 1.90 days) compared with mice with local injection of MSCs (6.94 ± 6.78 days) or no MSCs (5.93 ± 4.55 days)