Cratons are commonly considered as stable parts of continents that can survive a long-term interaction with mantle convective instabilities, basal drag and plate tectonic processes. However, geochemical evidence, geophysical observations and numerical modeling question their long-term stability and suggest heterogeneous modification with possible partial destruction of cratonic lithosphere. Cratonic modification may be identified either from a significant reduction in lithospheric thickness or from densification of cratonic lithospheric mantle, e.g. through melt-metasomatism. Both characteristics can be identified through geophysical modeling, which allows to constrain thermo-compositional heterogeneity of cratonic lithospheric mantle. The examples from the East European, Greenland, Siberian, North China, and Kaapvaal-Zimbabwe cratons, as well as the Arabian Shield, Tibet and Antarctica demonstrate various degrees of lithosphere reworking by mantle convective instabilities and plate tectonics processes. The observed broad variability in the present-day cratonic lithosphere structure precludes unique interpretations of past interactions of the cratons with mantle convection and plate tectonics processes, and indicates the existence of various types and multiple phases of such interactions, controlled by lithosphere rheology.