In order to reflect the features of wind farm power variation, this paper puts forward a method for regional wind power typical scenarios’ selection based on hierarchical clustering algorithm. Firstly, it uses the hierarchical clustering algorithm to cluster the wind power output samples and gains a clustering tree to reflect the similarity relation between samples. Then, in order to improve the quality of wind power typical scenarios’ selection, the sum of squares of deviations is used to describe the difference between interclass samples, which is regarded as a basis to determine the number of clusters, and it realizes the samples’ effective division. Finally, by using the real wind power output data in a certain region, it verifies the method’s rationality, and selects regional wind power typical scenarios meeting the requirements of peak regulation and reactive power configuration. This work is supported by National Natural Science Foundation of China (No. 51190103).