The principles and development of the stratigraphic approach for ident
ifying and studying traces of large old earthquakes are presented and
discussed. Large earthquakes can be detected in Quaternary sections, p
redominantly for the Late Pleistocene and Holocene, in two ways. First
ly, unusual deposits or facies may appear which are foreign to the lon
g-term depositional environment under study. Secondly, fracture and/or
fold deformations bearing traces of impulsive, instantaneous origin m
ay be detected. Experience in this work involves two different categor
ies with respect to tectonic and seismic activity areas in the Tien Sh
an and in the Crimea. The stratigraphic method opens new vistas in Qua
ternary geology and paleoseismology.