Helium (He) provides important information on chemical and isotopic mantle heterogeneities since it behaves as both incompatible and volatile element, and his recycling is limited during subduction pro- cesses. Nevertheless, the fundamental physical be- havior of He in mantle minerals (e.g., storage sites and diffusion mechanisms) remained poorly under- stood at high pressure and high temperature. As an incompatible element, He is preferentially stored in defects within the crystal structure in mantle rocks, such as point defects (i.e., Mg vacancies and intersti- tial sites), linear defects (i.e., dislocations), planar defects (i.e., grain boundaries), and 3-dimensions defects (i.e., pores and inclusions). Recent experi- mental studies were able to constrain He storage in polycrystalline olivine, settling that He is preferen- tially stored in grain boundaries rather than in point defects within the crystal lattice ([1], [2]). It implies that ~22% of He amount is stored in grain bounda- ries at typical mantle grain sizes, inducing a signifi- cant enhancement of bulk diffusivities compared to lattice diffusivities. Nevertheless, He storage and transport in planar defects is still poorly understood as well as the implications of deformation processes also remain to be determined. Since incompatible elements are preferentially stored along dislocations in zircon ([3]), the same behavior is expected to oc- cur for He in mantle mineral lattice. In this study, the implications of deformation processes on He storage and transport have been tested in deformed fine-grained synthetic polycrys- talline forsterite. The starting material consisted in sintered forsterite aggregates with a grain size of ~3 μm. Samples were then deformed in axial compres- sion at 300 MPa and 950, 1050 or 1200 °C using a Paterson press. Three deformed samples and one undeformed sample were subsequently doped in presence of He source (i.e., uraninite from Mis- tamisk, Canada) at 1 GPa and 1120 °C in a piston cylinder apparatus. Helium was then analyzed by coupling a cycled step heating protocol with a noble gas mass spectrometer. This method permits to de- termine He diffusivity for each temperature step. Additionaly, SEM and TEM analyses were per- formed on pre-doped samples to constrain textures and microstructures. Our results show complex diffusive behaviors with diffusivities that cannot be fitted by a single linear regression. Thus, a F-test has been performed on each individual step heating cycle showing that diffusivities can be fitted by several linear regres- sions. It highlights the competition between different diffusion mechanisms related to different He storage sites (Mg vacancies, interstitial sites, dislocations, and grain boundaries). Activation energy (Ea) and pre-exponential factor (D0) for He grain boundary diffusion have been refined from previous studies (Ea = 36 ± 9 kJ·mol–1 and D0 = 10–10.57 ± 0.58 m2 ·s –1 ), while those of He diffusion in intersitial sites (Ea = 89 ± 7 kJ·mol–1 and D0 = 10–8.95 ± 1.16 m2 ·s –1 ) and Mg vacancies (Ea = 173 ± 14 kJ·mol–1 and D0 = 10–5.07 ± 1.25 m2 ·s –1 ) are obtained from our results and litera- ture data. A last set of diffusion parameters included between those of He diffusion in grain boundaries and those in interstitials are interpreted as corre- sponding to He diffusion along dislocations (Ea = 56 ± 1 kJ·mol–1 and D0 = 10–9.97 ± 0.37 m2 ·s –1 ). By applying these results to mantle rocks with the highest dislocation density and millimetric grain size, a maximum He fraction of only 1.2% can be stored along dislocations. This value is well below the He fraction of 22%, which can be stored in grain boundaries at typical mantle grain sizes. Moreover, bulk diffusivities are affected by the presence of He in grain boundaries but the He amount stored along dislocations is too small to significantly modified bulk lattice diffusivities, regardless of the dislocation density. It implies that deformation processes could only increase He storage capacity and mobility in mantle rocks by reducing grain size (via dynamic recrystallisation). This process can implicate an in- crease of bulk concentrations of the deformed peri- dotites upon equilibration with nearby undeformed (or less deformed) peridotites.