Media shows up in a minority of dreams, in either sleep stage
The most direct evidence comes from a scoping review of 29 experiments that exposed people to moving visual media before sleep and then collected dream reports [1]. The share of reports containing something related to the pre-sleep stimulus was modest: 3% to 43% for REM dream reports and 4% to 30% for NREM sleep mentation reports [1]. In plain terms, even in the best cases most dreams did not incorporate the media, and in the weakest cases incorporation was rare.
The two ranges overlap almost completely, which matters for the REM-versus-NREM question. A study landing at 30% incorporation in NREM would sit above a study landing at 3% in REM, so the stage label alone does not tell you how much media will appear. The review describes the effects overall as modest yet varied, and it emphasizes large methodological heterogeneity across studies rather than a clean stage difference [1].
Why REM and NREM reports can look so similar
One reason the stage comparison is murky is that dream reports are filtered through memory and awakening. Reports after NREM awakenings tend to be shorter, less emotional, and less visually vivid than REM reports, but NREM awakenings still produce dream reports [4]. If NREM reports are thinner on average, a pre-sleep image has fewer chances to surface in them, which could compress the NREM incorporation rate without meaning that NREM sleep is less responsive to media.
The timing and quality of awakening also shape what gets reported. Different patterns of neural activity before awakening may affect whether a dream memory is stored or retrieved at all, and sleep inertia after awakening from different stages may affect the content and complexity of the report [3]. So a lower incorporation rate in one stage could reflect reporting conditions rather than a difference in how the sleeping brain treats the media.
Home dream reports sit in the same modest band
Studies that collected dream reports at home, rather than waking people in a lab, found incorporation rates between 11% and 35% [1]. That band sits inside the REM and NREM ranges, which is consistent with the idea that pre-sleep media influence is modest across settings and not sharply stage-specific. It also means the lab-versus-home distinction does not rescue a strong REM advantage.
What remains genuinely open is why incorporation is low in the first place. The review notes the general difficulty of influencing dream content with pre-sleep visual media and points to methodological gaps across studies [1]. The stage comparison is further complicated because REM sleep can occur without any dream report—about 20% of awakenings after a REM episode yield no report—so absence of media in a report is not proof that the sleeping brain ignored it [4].
Sources used in this answer
Impact of Pre-Sleep Visual Media Exposure on Dreams: A Scoping Review
Across 29 experimental studies, pre-sleep visual media was incorporated into 3% to 43% of REM dream reports, 4% to 30% of NREM sleep mentation reports, and 11% to 35% of home dream reports, with modest and varied effects overall.
Dreaming in sleep disorders: a comparative analysis of REM and NREM parasomnias
Patients with REM sleep behavior disorder showed higher dream recall and complexity and more fight behavior than patients with NREM parasomnias, and speech graph analysis distinguished the two groups, but the study did not examine pre-sleep visual media.
Dreaming and its Mechanisms in REM and NREM Sleep
Dreams are reported from both REM and lighter NREM sleep, and differences in pre-awakening neural activity and sleep inertia may affect whether and how dream content is stored, retrieved, and reported.
Characteristics of REM and NREM Dreams
NREM dream reports tend to be shorter, less emotional, and less visually vivid than REM reports, yet dream reports can occur without REM and about 20% of REM awakenings produce no dream report.
In the Arms of Morpheus: Recent Advances in Dreaming and in Other Sleep-Related Metacognitions
This compilation covers basic and clinical research on dreaming, including pre-sleep visual media exposure, but does not provide explicit evidence comparing REM and NREM dream reports.