Abstract: In this paper, the application scope of fixed threshold and percentile threshold index is compared during extreme temperature events. Also, the determination of percentile thresholds for defining extreme temperature indices in studies of longterm change of extreme temperature events is briefly summarized, and the differences of data and calculation methods used in different studies are compared. The effect of reference-period selection on percentile threshold values and the results of analyses are discussed. It is clear from the overview that both advantages and disadvantages exist for different methods of percentile threshold determination. It is found that the long-term trends of extreme temperature events detected are almost the same when different methods of calculation are used. In order to enhance the significance of the statistical analysis of extreme events, trends and to compare the analyses among different studies and different regions, however, it is necessary to further improve the methods of data selection, data processing, threshold calculation methods and the usage of climate baseline period. It is advisable to consistently use the uniform methods for determining percentile thresholds.