Comparative study of high order statistics estimators Arnaud Martin & Ali Mansour ENSIETA, E312 2 rue Francois Verny 29806 Brest, codex 9, France Abstract: To characterize and process a signal, many high order statistics are used by the signal processing researchers. Specific features of the data (temporal, stationary) and real time applications require the development of new estimators. In this paper, we study some estimators of high order moment and cumulant using adapted to different kind of signals. Key words: Estimators, high order statistic, moment, stimulant, temporal and non-stationary data.