If the technological progresses of additive manufacturing have enabled to achieve outstanding components with complex geometries, without adversely affecting their core mechanical and structural properties, their degraded surface state remains a critical issue which constitutes a key obstacle for its broad deployment. Indeed, surfaces of parts elaborated by this technic are characterized by a high roughness (Ra up to 40 µm), a strong texture (directly linked to the process and the manufacturing parameters) and to the presence of potentially detachable unmelted particles. Evaluating and developing relevant finishing post-treatments is the aim of the AFTER-ALM project (AFTER-Additive Layer Manufacturing) managed by the IRT M2P (Metz, France) which brings together industrial partners with major OEM such as Airbus, Safran, Liebherr, Collins Aerospace, Stelia, NavalGroup and ArianeGroup. The project also concerns companies specialized in additive manufacturing production (LISI Aerospace, WeAreAerospace, Spartacus 3D, Add-up, Volum-e), surface treatment applicators (Galvanoplast, Chrome Dur Industriel, DEC SA, Satys, GIT), technological partners (Micronics, ABC Swisstech, Beckmann Institute, IREPA laser, R&D Nano, INVENTEC, AQUARESE) and academics ones (UTINAM, CIRIMAT, ENISE).The work presented here focuses on development of an electrofinishing process dedicated to 316 L stainless steel developed by SLM. A study, performed at laboratory scale, allowed to characterize the electrochemical behavior of raw substrates (produced according to different laser scan strategies) was the opportunity to define the bests operating conditions for the levelling (electrolyte composition, temperature, electrical parameters U/I, treatment duration...) with acceptable dissolution rates (around 5 µm/min). The transposition to a pilot unit for the treatment of decametric sizes workpieces (flat or more complex specimens with variable grooves, tubes) requires a precise recalibration. Difficulties are essentially due to the high roughness of the SLM substrates (Ra ⋍30 µm, Rz ⋍ 200 µm), but also to issues related to the process scale-up such as current lines distribution which induces an inhomogeneous dissolution. It is possible to obtain very good results in roughness decrease, but to the detriment of the dimensional characteristics.To meet the dual objective of roughness reduction combined with respect for the geometrical integrity, the use of pulsed potentials has been shown to be highly effective. Tested at laboratory scale and then extended to the pilot scale, a 90% decrease of the roughness was measured while preserving as much as possible the geometries of all samples. These good performances were explained by differences in the mechanisms involved with constant or pulsed potential, as highlighted by transient curves study (figure 1) and spectral power density processing of roughness data. Figure 1