We consider materials which can be described by plasticity laws exhibiting nonlinearkinematic and nonlinear isotropic hardening effects. The aim is to show that the materialparameters governing the constitutive behavior may be determined from data obtained byspherical indentation. Note that only the measurable global quantities (load and indentationdepth) should be utilized, which are available, e.g. from depth-sensing indentation tests. For thisgoal use is made of the method of neural networks. The developed neural networks apply also tothe case of pure kinematic as well as pure isotropic hardening. Moreover it is shown how amonotonic strain–stress curve can be assigned to the spherical indentation test.