Transformer is one of the most expensive and critical equipment in power industry. The research on condition monitoring and fault diagnosis of transformer is significant for the stability of power system. However, transformer diagnosis involves many factors, so it is very difficult to evaluate transformer operation state effectively and accurately by single diagnosis method. In this paper, we present a novel interactive visualization model of transformer diagnosis based on multiple views.This model imitates multi-angle analysis by combining multi-source information including operation information, experiment information and environmental information through multi-view collaborative visualization. Data are displayed in different dimensional and a variety of interaction technology are used so that analysts can access the original data from multiple perspectives. With the help of visual perception of human eyes and intelligent cognitive ability of human brain, users can find useful information from large-scale, high-dimensional, or even incomplete data to make effective decisions.