Abstract: China has been working to alleviate since its reformation, and has made certain success. However, the poverty-stricken populations in China distributes in rural in mountainous areas for a long time. It is difficult to get rid of poverty in these lagging areas. A comprehensive understanding of the poverty situation in mountainous areas and effective techniques to reduce it have become important tasks for building a healthy society. We chose six typical villages in Taihang Mountain, Beiye Township, Pingshan County, Hebei Province based on elevation and geographical characteristics as the objectives. A question-based survey was conducted to obtain relative data. The evaluation index system was established with three dimensions-income, capital and living standard-and 9 indexes. The A-F (Alkire-Foste) method was used to analyze the results of multidimensional poverty measurement. The results showed that:1) the single-dimensional poverty situation in Beiye Township was still serious, and the causes of poverty were significantly different in all dimensions. Poverty mainly manifested as the absence of capital dimension. 2) The degree of multi-dimensional poverty in Beiye Township was deep. More than 90 percent of poor households lived in poverty. With an increase in dimension k, the multi-dimensional poverty index decreased. 3) The contribution of each index to the multi-dimensional poverty of Beiye Township was significantly different. The indexes of per capita cultivated land area, education, skill training and health facilities all contributed more than 15% and changed little when dimension k increased. 4) The lack of education was an internal factor that led to poverty. Compared to families with education, families without education had lower monthly income, single income source, and larger family size. In view of the above conclusions, this paper put forward the suggestions of using a hierarchical optimization mode to organize rural residential areas so as to integrate and optimize educational resources and help mountainous areas eliminate poverty.