Experiments have consistently shown that metallic materials display strong size effects at the micron scale, where the peculiar phenomenon “smaller is stronger” appears. Considering the exponentially increasing trend towards miniaturization, correct modeling of such effects has become inescapable in various high-technology fields, including microbotics, microelectronics, micromedicine, etc. As a result, a strong scientific effort has been devoted to the subject and numerous size-dependent theories have been proposed in recent years. Among them, strain gradient plasticity (SGP) theories, which can be seen as an extension of conventional plasticity to small scales, can particularly be cited. Including internal length scale(s), these theories are capable of predicting plastic deformation gradients, which correlate with size effects as experimentally observed and numerically predicted using dislocation mechanics. Thanks to their capabilities in capturing size effects, SGP theories have become increasingly used for the prediction of the size-dependent behavior of materials at the micron scale. However, despite the significant progress made on these theories, they still suffer from some fundamental issues, so far preventing their application to real engineering problems. The present project is proposed to address one of the most challenging of these issues: the physical nature of elastic gaps which are numerically observed using major of such theories. Almost all existing thermodynamically-consistent SGP theories including higher-order dissipation predict elastic gaps (delay in plastic flow) under certain non-proportional loading conditions. Nevertheless, to date, there is no experimental evidence nor is there small-scale numerical confirmation proving the existence of such gaps in reality. This represents a major source of confusion and uncertainty, preventing the development of robust SGP models that can be applied in a real industrial context. In the absence of works studying these gaps from a physical point of view, the scientific investment on SGP theories has reached a bifurcation point making the scientific community divided into those who consider elastic gaps as “unknown” size effects (then possibly physical) and those who see no physical reasons for their occurrence in reality, at least within a single-crystal, as they reflect an instantaneous finite change in the higher-order stresses. This project aims at clearing-up this ambiguity and at providing the compelling answer to the question "are elastic gaps physical?". To achieve its goals, the project will create a multi-disciplinary synergy between original small-scale experiments, implying, for the first time, non-proportional loading conditions, and extensive discrete-dislocation-based simulations. Results of the experimental and numerical investigations will then be considered to review major existing SGP theories, with the aim of developing the first single- and poly-crystal SGP models that are robust enough for real engineering applications. As an application, the proposed models will finally be used to investigate one of the most challenging small-scale problems: impact of size effects on the formability of ultra-thin sheet metals, which show an ever-growing use in various engineering fields.