In this article we present a critique of the sensory conflict theory of motion sickness. We discuss three forms of sensory conflict that are believed to exist: input conflict, output conflict, and expectancy violation. We argue that any concept of sensory conflict necessarily entails the violation of internal expectations and, hence, that the three forms of conflict are logically identical. In the motion sickness literature, it is implicit that the expected pattern is one in which the inputs of the visual, vestibular, and somatosensory systems are redundant. We argue that an expectation of redundancy cannot plausibly result from interaction with the environment. We review data indicating that nonredundant patterns of intermodal stimulation lead to adaptive changes in the control of behavior; this implies that the calculation of nauseogenic conflict must be unrelated to the control of behavior. Moreover, we argue that no single pattern of sensory stimulation, redundant or otherwise, can serve as a reliable standard or expectation for the calculation of sensory conflict. We conclude that there is no principled basis on which the conflict theory can distinguish between nauseogenic and nonnauseogenic situations. It follows that the concept of sensory conflict cannot provide a theoretical explanation for the existence of motion sickness. We review data suggesting an important role for the control of behavior in the etiology of motion sickness. This suggests a new basis for understanding the phenomena of motion sickness. © 1991, Taylor & Francis Group, LLC. All rights reserved.