Dry-hot wind day is one of the main argometeorological disasters which impact the development and yield of winter wheat in the North China Plain (NCP). Using daily meteorological data of 48 weather stations from the year of 1961 to 2008 in the NCP, the dry-hot windy days, that include two types, type one is high temperature and low humidity and type two is immature death after rain, are analyzed. Then by using dry-hot windy days and Cluster Analysis of Statistic Test (CAST), which is a newly developed regionalization method, the dry-hot windy days regionalization is carried out in the NCP. Results show that spatially speaking, both types of dry-hot windy days all present higher in the middle and lower in the south and north; temporally speaking, type one is slightly decrease in the last nearly fifty years, except the individual stations in the south part of the plain. Although type two of dry-hot windy days is also generally decrease in the past, however, the exceptional stations are higher than type one. Results of CAST indicate that for the type one, five regions are divided and for the type two, four regions are divided. Those regionalization results are reasonably reflecting the spatial and temporal changes of dry-hot windy days in those load centers, and CAST is more powerful than conventional regionalization methods. The results could provide the scientific and practical base for developing the monitor and forecast system of dry-hot wind days, and is also able to provide realistic instructions for the decision making on the agricultural managements of the local farmers.