Managing projects in an organization, especially a project-oriented organization, is a challenging task. Project data has a large volume and is complex to manage. It is different from managing a single project, because one needs to integrate and synthesize information from multiple projects and multiple perspectives for high-level strategic business decisions, such as aligning projects with business objectives, balancing investment and expected return, and allocating resources. Current methods and tools either do not well integrate multiple aspects or are not intuitive and easy to use for managers and executives. In this dissertation project, a multidimensional and visual exploration approach was designed and evaluated to provide a unique and intuitive option to support decision making in project portfolio management. The research followed a general design science research methodology involving phases of awareness of problem, suggestion, development, evaluation and conclusion. The approach was implemented into a software system using a prototyping method and was evaluated through user interviews. The evaluation result demonstrates the utility and ease-of-use of the approach, and confirms design objectives. The research brings a new perspective and provides a new decision support tool for project portfolio management. It also contributes to the design knowledge of visual exploration systems for business portfolio management by theorizing the system.