Energy management systems provide an opportunity to collect vast amounts of building-related data. The data contain abundant knowledge about the interactions between a building’s energy consumption and the influencing factors. It is highly desirable that the hidden knowledge can be extracted from the data in order to help improve building energy performance. However, the data are rarely translated into useful knowledge due to their complexity and a lack of effective data analysis techniques. This paper first conducts a comprehensive review of the commonly used data analysis methods applied to building-related data. Both the strengths and weaknesses of each method are discussed. Then, the critical analysis of the previous solutions to three fundamental problems of building energy performance improvement that remain significant barriers is performed. Considering the limitations of those commonly used data analysis methods, data mining techniques are proposed as a primary tool to analyze building-related data. Moreover, a data analysis process and a data mining framework are proposed that enable building-related data to be analyzed more efficiently. The process refers to a series of sequential steps in analyzing data. The framework includes different data mining techniques and algorithms, from which a set of efficient data analysis methodologies can be developed. The applications of the process and framework to two sets of collected data demonstrate their applicability and abilities to extract useful knowledge. Particularly, four data analysis methodologies were developed to solve the three problems. For demonstration purposes, these methodologies were applied to the collected data. These methodologies are introduced in the published papers and are summarized in this paper. More extensive investigations will be performed in order to further evaluate the effectiveness of the framework.