The literature on aspirations and risky organizational change has documented a wide range of empirical patterns and relied on a mix of theories to explain them, resulting in fragmented insights and limited predictive power. We develop a formal model to address these limitations. In our model, organizations evaluate projects based on reference-dependent utility functions. We demonstrate that the diverse patterns observed in the literature can be explained by a small set of behaviorally-plausible assumptions regarding the shape of the organization’s utility function, the characteristics of the projects being evaluated, and the noisiness of the evaluation process. Our model provides a parsimonious micro-foundation to the literature on aspirations and risky organizational change as well as many testable predictions regarding the propensity of risky organizational change.