Abstract: New advances in extremes statistics are reviewed in this paper, with emphases on application to extreme weather and climate event research as well as in large-scale engineering design. For example, the generalized Pareto distribution (GPD) based on the peak, the generalized extreme value distribution (GEV) based on the block maxima over threshold (POT) method, and the theoretical relationship of parameters between GPD and GEV; the quantile estimation of extreme precipitation model with given return periods; and the quantile estimation error of extreme value distribution models, and the multivariable extreme value distribution theory, and its application, etc.