INTRODUCTION: Mesenchymal stromal cells (MSCs) gained attention for their anti-inflammatory and trophic properties, with musculoskeletal diseases and osteoarthritis (OA) being among the most studied conditions. Alongside cells, their released factors and extracellular vesicles (EVs), overall termed "secretome", are actively sifted being envisioned as the main therapeutic actors. In addition to standard supplementation given by foetal bovine serum (FBS) or human platelet lysate (hPL), new good manufacturing practice (GMP)-compliant serum/xeno (S/X)-free media formulations have been proposed, although their influence on MSCs phenotype and potential is scarcely described. The aim of this study is therefore to evaluate, in the OA context, the differences in secretome composition and potential after adipose-MSCs (ASCs) cultivation in both standard (FBS and hPL) and two next generation (S/X) GMP-ready supplements. METHODS: Immunophenotype and secretory ability at soluble protein and EV-related levels, including embedded miRNAs, were analysed in the secretomes by means of flow cytometry, nanoparticle tracking analysis, high throughput ELISA and qRT-PCR arrays. Secretomes effect was tested in RESULTS: Within a conserved molecular signature, a divergent fingerprint emerged for ASCs' secretomes collected after expansion in standard FBS/hPL or next-generation S/X formulations. Regarding soluble factors, a less protective feature for those in the secretome collected after ASCs were cultured in S/X media emerged. Moreover, the overall message for EV-miRNAs was characterized by a preponderance of protective signals in FBS and hPL conditions in a context of general safeguard given by ASCs released molecules. This dichotomy was reflected on secretomes' potential CONCLUSIONS: These data open the question about the implications from using new media for MSCs expansion for clinical application. Although the undeniable advantages for GMP compliant processes, this study results suggest that new media formulations would deserve a deep characterization to drive the choice of the most effective one tailored to each specific application.