In this research we develop, operationalize, and empirically test a model for explaining/predicting the satisfaction of customers with Internet-based services at different stages of adoption. We argue and empirically demonstrate the need to consider the evolutionary nature of satisfaction and the variability of its determinants. Our model identifies desire disconfirmation, expectation disconfirmation, and perceived performance as the main determinants of satisfaction and differentiates between satisfaction at adoption of Internet-based services and satisfaction in the post-adoption stage. Our empirical results show that desires and expectations are both important comparison standards that need to be considered simultaneously in explaining satisfaction at adoption. The role of desires, however, diminishes significantly in the post-adoption stage. The results also show no significant relationship between post-adoption satisfaction and satisfaction at adoption. The paper presents the theoretical foundation of the proposed model and discusses the implications of the empirical results.