Playing a significant role in production process of tea enterprise, tea blending has advantages in improving the quality of tea, expanding resource of tea, and getting higher profits. Traditional manual blending method not merely wastes time and energy but also has difficulties in optimizing blending programs. In this paper, tea blending problem is modeled as spatial clustering based on multi-dimensional hierarchy space. Tea attributes such as varieties of tea bush, process crafts and producing areas are modeled into hierarchy space with tree structure. Every node of these trees is present a value of tea attributes. So all the data tuples are mapped as points in multi-dimensional hierarchy space. And we define a similarity-based measure criterion in multi-dimensional conceptual layered space to present an agglomerative hierarchical spatial clustering based algorithm to work out the optimal blending program. Meanwhile, Dewey code is introduced to increase resolution efficiency. Dewey code is used presented the points in hierarchy space, and the codes of every point can be adopted to compute the similarity of every two points in the measure criterion. Two clustering algorithms have been proposed. We agglomerate two points which are the closest to each other according to the similarity measure criterion in hierarchy space by algorithm AGHC. In DIHC, the whole dataset is divided into small parts through K-medios until there are N clusters and N is set by users. Tea attributes are quantified and the process of tea blending is standardized. This study enable tea blending to get rid of human experience, instead of intelligent approach. The results of tea blending in this article can be accurate due to the strict consolidation doctrine. That means we will be rational when we choose which tea to be blended in practice to satisfy customers’ demands. Appropriate tea will be blended by several kinds of tea with the study in this article. Finally, experiments on real data set demonstrate the solution of tea blending proposed in this paper is greatly improving work efficiency and economic benefits.