Abstract It was shown experimentally that changing the indenter load P and the indentation size does not change the total deformation during indentation ( e t ≈ const.), and that fracture does not influence the scale effect. For this reason the physical nature of the scale effect is revealed better by nanoindentation and under more “clean” conditions than in uniaxial deformation tests. The indentation size effect (ISE) is revealed as the change of mechanical properties determined by indentation. It was shown that reduction of the indent size leads to both increasing hardness and decreasing plasticity, determined by indentation. The phenomenological approach to the ISE (in which the power dependence of the indenter load P on the indenter displacement of h is used) made it possible to describe the dependence of nanohardness H ( P ) and H ( h ) by simple equations. Nanohardness was determined for 21 different crystals, and parameters that enabled the size dependence of H for these crystals to be calculated were determined. It is proposed to determine nanohardness at h = const. instead of P = const. and to recalculate H using our equations for fixed values of h f = 1000 nm for metals and h f = 100 nm for hard materials. The use of the developed technique makes it possible to compare results of nanohardness tests from different sources for different indenter loads.