Introduction. Inner speech, also referred to as inner voice or verbal imagery/thought, is the experience of speaking to oneself in the mind’s ear, including the ability to simulate future conversations or replay past conversations with relative ease. While this is a fairly common ability, it is also a variable one, with some individuals reporting vivid and frequent verbal thoughts (hyperendophasia), while others report little to no experience of inner speech (anendophasia; Nedergaard & Lupyan, 2024; Zeman, 2024). Previous research has suggested that engaging in inner speech is accompanied by subtle activity in articulatory muscles involved in overt speech production, suggesting a close relationship between inner speech and the articulatory motor system (Fernyhough & Borghi, 2023; Livesay et al., 1996; Moffatt et al., 2020; Nalborczyk et al., 2020). Additionally, work on overt reading has shown right lateralisation of articulatory muscle activity, raising the question of whether similar lateralisation patterns are also present during inner speech (Nicholls & Searle, 2006; Wolf & Goodale, 1987). Finally, reduced vividness of verbal imagery may be associated with reduced engagement of the speech motor system, as reflected in articulatory muscle activity during engagement in verbal thoughts and related verbal tasks. Methods. In this study, we investigated the articulatory correlates of inner speech by recording surface electromyography (EMG) from the speech muscles of 26 right-handed participants so far during four conditions: visual rest, silent reading, verbal thinking, and overt reading aloud. The reading and thinking sessions involved two chapters of The Little Prince, with reading sessions being self-paced, and visual rest and verbal thinking sessions being 3 minutes long. Participants completed two sessions of each condition. EMG activity was recorded bilaterally from five facial muscle groups associated with speech articulation: inner superior orbicularis oris, outer superior orbicularis oris, inner inferior orbicularis oris, outer inferior orbicularis oris, and zygomaticus major (Moffatt et al., 2020; Nalborczyk et al., 2020; van Boxtel, 2001). Results. Preliminary analyses of EMG signal amplitude revealed reliable differences in articulatory muscle activity across conditions. Crucially, inner speech elicited lower EMG activity than overt speech, but higher activity than visual rest. These effects are observed both in terms of overall signal magnitude and the temporal evolution of the signal over the course of each task, and are mapped across all ten EMG channels. We further explored facial lateralisation of articulatory activity across conditions, and preliminary findings suggest that articulatory potentials tend to show right lateralisation during overt speech, whereas a relative shift towards left lateralisation was observed during verbal thinking. Finally, we discuss how reduced vividness of verbal imagery influences articulatory muscle activation across conditions. Conclusion. These findings support the view that inner speech engages the speech motor system in a graded manner, intermediate between overt speech and non-verbal resting states. More broadly, our preliminary findings suggest that different forms of internal and external language engagement are associated with distinct spatiotemporal patterns of articulatory activity, and might be modulated by the vividness of verbal imagery.