Abstract Introduction Prolonged Exposure (PE) therapy produces therapeutic fear extinction via imaginal exposure to trauma memories. However, traumatic events that occurred in the distant past and the associated memories may become distorted or habituated. Posttraumatic nightmares are more recent, potentially salient, and may better support extinction learning. Physiological responses to imagery of a trauma and nightmare related to this trauma were compared to each other and to neutral imagery. Methods Twelve participants (mean age=26.16, 11 female) with PTSD (mean CAPS-5=27.83) and frequent trauma-related nightmares wrote accounts of their trauma. Participants then completed a 14-day sleep-monitoring period with diaries, actigraphy and two nights of ambulatory PSG. Participants narrated a nightmare report into an audio recorder when awoken by a nightmare or when recalled upon awakening. Two pairs of short narratives were created from the written account of the trauma and recording of a nightmare most similar to the trauma. These narratives (scripts) were audio-recorded by an investigator. Participants then underwent two script-driven imagery (SDI) sessions, one hour apart, during which they listened to either their two trauma-memory or their two nightmare-memory scripts (counterbalanced across participants) with 3 interspersed neutral scripts. Each script in an SDI session included baseline, listening, and imagery periods (approximately 30 sec apiece). Skin conductance (SC), heart rate (HR), and corrugator electromyography (EMG) biosignals were continuously recorded throughout each SDI session. For each script, HR, SC, and EMG means during the baseline period were subtracted from their respective imagery-period means. These difference scores were square-root transformed and analyzed by ANOVA with Type (trauma vs. nightmare) and Valence (trauma/nightmare vs. neutral) factors. Results Biosignals from scripts of both Types (trauma and nightmare) significantly exceeded those from their respective neutral scripts [HR:F(1,11)=23.42, p=0.0005; SC:F(1,11)=9.53, p=0.01; EMG:F(1,10)=8.0, p=0.018]. However, biosignals from trauma and nightmare scripts did not differ (p’s>0.39) nor did the Type x Valence interactions (p’s>0.10). Conclusion Physiological reactivity during imagery of a trauma memory and a trauma-related nightmare both significantly exceeded reactivity to neutral scenarios. Nightmare-memory and trauma-memory imagery produced similar reactivity. Thus, imagery of nightmares have potential utility as alternative PE stimuli. Support (If Any) This project was supported by NIMH grant 1R21MH121832-01A1 to E.P.S.