The objectives of this study were to use non-equilibrium gravitational field-flow fractionation (GrFFF), an immunotag-less method of sorting mesenchymal stem cells (MSCs), to sort equine muscle tissue-derived mesenchymal stem cells (MMSCs) and bone marrow-derived mesenchymal stem cells (BMSC) into subpopulations and to carry out assays in order to compare their osteogenic capabilities. Cells from 1 young adult horse were isolated from left semitendinosus muscle tissue and from bone marrow aspirates of the fourth and fifth sternebrae. Aliquots of 800 3 103 MSCs from each tissue source were sorted into 5 fractions using non-equilibrium GrFFF (GrFFF proprietary system). Pooled fractions were cultured and expanded for use in osteogenic assays, including flow cytometry, histochemistry, bone nodule assays, and real-time quantitative polymerase chain reaction (qPCR) for gene expression of osteocalcin (OCN), RUNX2, and osterix. Equine MMSCs and BMSCs were consistently sorted into 5 fractions that remained viable for use in further osteogenic assays. Statistical analysis confirmed strongly significant upregulation of OCN, RUNX2, and osterix for the BMSC fraction 4 with P , 0.00001. Flow cytometry revealed different cell size and granularity for BMSC fraction 4 and MMSC fraction 2 compared to unsorted controls and other fractions. Histochemisty and bone nodule assays revealed positive staining nodules without differences in average nodule area, perimeter, or stain intensity between tissues or fractions. As there are different subpopulations of MSCs with different osteogenic capacities within equine muscle- and bone marrow-derived sources, these differences must be taken into account when using equine stem cell therapy to induce bone healing in veterinary medicine.