Accurate determination of the tensile stress–strain response of structural materials typically relies on standardized tensile specimens. In many practical cases—such as limited material availability or component-scale constraints—only sub-size geometries can be produced. However, size-dependent effects often cause tensile curves obtained from sub-size specimens to deviate from those of standardized ones [1], hindering quantitative comparison beyond qualitative trends. This study presents a Finite Element (FE)–based inverse methodology capable of reconstructing the stress–strain curve of a standardized specimen using experimental tensile data from non-standard specimens only. Tensile tests were conducted on flat standard specimens (100 mm total length) and on sub-size geometries (46 mm total length) for two metallic materials with distinct mechanical responses: Aluminum 6061-T6 and 304L stainless steel. Corresponding FE models were generated for each geometry and material. A ductile damage formulation was incorporated to capture strain localization, necking, and final fracture [2]. Material-specific damage and hardening parameters were calibrated using experimental data from sub-size specimens and corresponding FE simulations. These calibrated parameters were subsequently applied to FE simulations of standard tensile geometries. The numerically reconstructed stress–strain curves showed good agreement with experimental measurements obtained from actual standard-specimen tests, demonstrating that the full equivalent stress–strain response can be retrieved from sub-size data alone. The proposed experimental–computational framework is general and currently being extended to additional materials and geometries. It offers a practical route for material characterization when fabrication of standard specimens is infeasible, and enables quantitative comparison of results obtained from diverse specimen sizes.