There is considerable evidence to suggest that consumer dissatisfaction with self-service technologies is widespread. However, there has been little conceptual or empirical scrutiny of the likelihood that consumers will complain to an organization (likelihood of voice) in this context. This study contributes to the service domain by testing empirically a model of the antecedents of consumers' likelihood of voice in unsatisfactory encounters with self-service technologies. A model is tested that combines established antecedents of voice, such as likelihood of voice success, and those that have not yet been considered, including self-service technology powerlessness and need to vent. The results support the proposed model in general. Theoretical and managerial implications of the findings are discussed.