Candidate cell sources for vocal fold scar treatment include mesenchymal stromal cells from bone marrow (BM-MSC) and adipose tissue (AT-MSC). The mechanosensitivity of MSC can alter highly relevant aspects of their behavior, yet virtually nothing is known about how MSC might respond to the dynamic mechanical environment of the larynx. Our objective was to evaluate MSC as a potential cell source for vocal fold tissue engineering using a mechanically relevant in vitro model (Study 1) and in vivo model (Study 2). In Study 1, a bioreactor capable of reproducing the vocal mechanoenvironment was used to evaluate the genomic similarity between MSC and vocal fold fibroblasts (VFF) following vibratory strain. Posterior probabilities for each of the 53,617 microarray transcripts fitting into specific expression patterns were calculated. VFF shared more Gene Ontology (GO) wound healing terms with BM-MSC than with AT-MSC. For example, vibrated VFF and BM-MSC shared similar capacities for glycosaminoglycan and proteoglycan activity, which could favorably effect collagen remodeling in vocal scar. Analysis of GO terms relating to MSC differentiation toward osteogenic, adipogenic, and chondrogenic lineages revealed that BM-MSC expressed fewer osteogenesis terms in the vibrated and scaffold conditions as compared to the polystyrene condition. This was consistent with our hypothesis that tissue-relevant stress encourages tissue-relevant cell differentiation. Lastly, mechanotransductive mechanisms underlying phonatory biology were preliminarily discussed with regard to microarray data, with integrin and lamellipodia activity distinguishing vibrated from non-vibrated cells. Microarray data was confirmed with real-time PCR. Immunostaining for Ki67 (proliferation) was positive in 15- 23% of cells, and cleaved caspase 3 (apoptosis) was positive in 2-7% of cells, and neither varied with cell type or mechanical condition. In Study 2, mature rabbit vocal scars were treated with four variations of MSC therapy. Nine weeks following injection, vocal folds receiving BM-MSC without a hyaluronic acid hydrogel produced the highest levels of a desirable adipogenesis marker (lipoprotein lipase) and an undesirable osteogenesis marker (osteocalcin). Osteocalcin was expressed 452-fold lower than in bone tissue, suggesting a clinically insignificant vocal fold level. As our field progresses toward clinical trials of cell-based vocal scar therapies, consideration of mechanotransductive and differentiation profiles is prudent.