Articular cartilage is a tissue characterized by its poor intrinsic capacity for self-repair. This tissue is frequently altered upon trauma or in osteoarthritis (OA), a degenerative disease that is currently incurable. Similar musculoskeletal disorders also affect horses and OA incurs considerable economic loss for the equine sector. In the view to develop new therapies for humans and horses, significant progress in tissue engineering has led to the emergence of new generations of cartilage therapy. Matrix-associated autologous chondrocyte implantation is an advanced 3D cell-based therapy that holds promise for cartilage repair. This study aims to improve the autologous chondrocyte implantation technique by using equine mesenchymal stem cells (MSCs) from bone marrow differentiated into chondrocytes that can be implanted in the chondral lesion. The optimized protocol relies on culture under hypoxia within type I/III collagen sponges. Here, we explored three parameters that influence MSC differentiation: culture times, growth factors and RNA interference strategies. Our results suggest first that an increase in culture time from 14 to 28 or 42 days lead to a sharp increase in the expression of chondrocyte markers, notably type II collagen (especially the IIB isoform), along with a concomitant decrease in HtrA1 expression. Nevertheless, the expression of type I collagen also increased with longer culture times. Second, regarding the growth factor cocktail, TGF-3 alone showed promising result but the previously tested association of BMP-2 and TGF-1 better limits the expression of type I collagen. Third, RNA interference targeting Col1a2 as well as Col1a1 mRNA led to a more significant knockdown, compared with a conventional strategy targeting Col1a1 alone. This chondrogenic differentiation strategy showed a strong increase in the Col2a1:Col1a1 mRNA ratio in the chondrocytes derived from equine bone marrow MSCs, this ratio being considered as an index of the functionality of cartilage. These data provide evidence of a more stable chondrocyte phenotype when combining Col1a1 and Col1a2 siRNAs associated to a longer culture time in the presence of BMP-2 and TGF-1, opening new opportunities for preclinical trials in the horse. In addition, because the horse is an excellent model for human articular cartilage disorders, the equine therapeutic approach developed here can also serve as a preclinical step for human medicine.