The identification of the potential benefits of a proposed information system project is a very challenging and important task. However, little research describes how this is actually carried out. The objective of this research was to answer the core question: how do organizations identify the expected benefits of proposed IS projects? To accomplish this objective, the authors studied 24 projects by interviewing 20 subjects from 13 organizations. Based on these interviews, a model of the benefit identification process was formulated showing it to be a set of loosely defined, overlapping, iterative activities. The study suggests that, despite great uncertainty about the potential benefits, proposers were compelled to use the benefit identification process to persuade the organization to be committed to the project. This effort to build commitment under great uncertainty may contribute to the overstatement or understatement of expected benefits. The study offers managers an understanding of the process and guidance for carrying it out. It also suggests that future researchers validate the model, use it to help identify best practices and test related hypotheses.