The problems of visualizing small-grain parallel programs are studied in this thesis. A visualization tool, called VISTA, was implemented using X-Windows. The goal of VISTA is to give effective visual feed­back to the user analyzing a program. To achieve thls goal, the tool summarizes a whole reduction tree in one window space using color. The user is able to see the reduction tree from different perspectives, such as PE, time, PE-time, and procedure views. VISTA also provides the user with several functions such as zoom-up, subtree display, and node inquiry to help analyze programs in more detail. The features and algorithms of VISTA are described in thls thesis. Examples are given of how the tool can be used to improve one's understanding of both small-grain parallel program characteristics.