When it comes to body movements in external space, people are experts in learning fine-grained voluntary control, for example, when manipulating tiny objects. Voluntarily controlling actions in the internal body (e.g., decreasing heart rate), however, is far more difficult and requires dedicated training, for example, in meditation or yoga. Not much is currently known about the learning mechanism underlying the acquisition of voluntary control over internal visceromotor actions (i.e., interoaction) or why it is so difficult compared to controlling our external somatomotor actions (i.e., exteroaction). We propose the ideomotor hypothesis of voluntary interoaction in this article, which asserts that voluntary exteroactions and interoactions are governed by the same general principle, namely, the anticipation of sensory feedback. We apply this hypothesis to two techniques that can be used to acquire voluntary control over interoactions, that is, autogenic training and biofeedback training. As the afferent signal we receive from interoaction (i.e., interoceptive signals from the internal body) is of lower sensory quality than the afferent signal that we receive from exteroaction (i.e., exteroceptive signals from the external environment), this hypothesis explains why learning to control interoactions is more difficult. We propose ways in which to test predictions from this hypothesis and show its theoretical value by comparing it to other frameworks in the literature. We hope that this work motivates future empirical studies directly examining voluntary interoaction and its clinical applications, such as autogenic and biofeedback training, and mind-body practices more generally. (PsycInfo Database Record (c) 2024 APA, all rights reserved).