From a neuroscience perspective, the reprocessing of information during sleep may give rise to dream content. Although dreaming is a common human experience, its electrophysiological correlates remain poorly understood. It has been proposed that dreaming does not depend on the cortex’s global oscillatory state but on localized activity in the parieto-occipital region. High-density EEG studies show that increased high-frequency (20-50 Hz) power in frontal and temporal areas during REM sleep is related to dream content. Importantly, these studies examined only spontaneous dreams, without prior incubation using learned information. Pilot work from our lab suggests that incubated dream content (linked to prior learned material) may reflect a global state rather than localized activity. We hypothesize that distinct EEG patterns will characterize spontaneous dreams versus dreams following new learning. To test this, participants will either learn a visuo-spatial location task in virtual reality or not, before 8 hours of sleep under polysomnography (EEG 10-20, 32 channels; EOG, EMG, ECG). In the last 4 hours, serial awakenings in phasic REM will collect dream reports. We will analyze the preceding segments for power spectral density, connectivity, and complexity (Lempel-Ziv, Shannon entropy). Here, we present the rationale and methodological design of Experiment 1 for discussion.