meta-Tetrakis(3-hydroxylphenyl)chlorin (mTHPC, Foscan®) is one of the most potent clinically approved photosensitizer (PS) for photodynamic therapy (PDT) and is characterized by high absorption in the deep red. mTHPC is partially aggregated in the bloodstream resulting in unfavorable biodistribution and relevant side effects. In order to improve the mTHPC tumor selectivity different attempts of nanovectorisation were conducted. Liposomal mTHPC (Foslip®) is monomeric in biological medium and has favorable pharmacokinetic properties. However, a rapid liposome destruction during circulation and rapid mTHPC release are obvious shortcomings. Alternatively, mTHPC nanovectorisation could be realized by extracellular vesicles (EVs). Derived from the cells, EVs possess a natural stability in bloodstream and ability to deliver cargo molecules into cells. This formulation is interesting for PDT due to the ability to encapsulate porphyrins and to redistribute them into cancer cells. Multicellular spheroid (MCTS) is a non-vascularized micro-tumor tridimensional model of cells. Compared to monolayer cells, MCTS show microenvironment similar to in vivosituation, making it as a suitable model to study PDT parameters. To determine whether mTHPC nanovectorisation by EVs could influence mTHPC distribution and photo-dynamic activity compared to liposomal and free mTHPC. Stability of mTHPC-EVs was performed by transmission electronic microscopy and NTA. Diffusion of PS in HT29 spheroids was studied by confocal microscopy and cell survival after irradiation by clonogenic assay. In vivo study was performed after HT29 subcutaneous xenografts. Membrane integrity of mTHPC-EVs was conserved in circulation in contrast to Foslip®, which is rapidly destroyed. In MCTS, mTHPC-EVs demonstrated 2 and 4-times higher photocytotoxicity compared to liposomal and free mTHPC. Indeed, 50% of EV-mediated cell photoinactivation was observed at 12 J.cm-2, while much less photokilling was noted with all mTHPC formulation. This result can be related to the increased mTHPC diffusion in MCTS, deeper than 100 μm, observed with mTHPC-EVs. In vivo preliminary results showed an improvement of PDT efficacy. In fact, 47 days were required to obtained 10-times initial tumor volume (10x V0) after intravenous injection of mTHPC-EVs compared to Foslip® and mTHPC where 10x V0 was already reached at 30 and 35 days respectively. Integration of mTHPC in EVs improves diffusion and photocytotoxicity in HT29 spheroids and xenografts. These results could be explained by an enhanced uptake of EVs by cancer cells and an increased stability in circulation. Overall, EVs should be considered as perspective nanocarriers for mTHPC-mediated PDT.