 The emergence of genetic data coupled to longitudinal electronic medical records, EMRs, offers the possibility of performing phenome-wide association studies, FEWAs. In FEWAs, the entire phenome can be divided into numerous phenotypic categories, according to the genetic architecture across phenotypes. Currently, statistical analyses for FEWAs are mostly unitivariate analyses, which test the association between one genetic variant and one phenotype at a time. In this paper, we derive a novel and powerful multivariate method for FEWAs. The proposed method involves three steps. Firstly, we use the bottom-up hierarchical clustering method, BHCM, to partition a large number of phenotypes into disjoint clusters within each phenotypic category. Secondly, the clustering linear combination method, CLCM, is applied to combine test statistics within each category based on the phenotypic clusters. Finally, we propose a new false discovery rate, FDR, control approach. We conduct extensive simulations to compare the performance of our method with that of other existing methods. The results show that our proposed method controls. This article was authored by Xiaoyu Liang, Xiuowei Chao, Zhou Yingxia, and others. We are article.tv, links in the description below.