Abstract Introduction Young adults, particularly those with histories of interpersonal trauma or stress, are more likely to experience to adverse psychosocial outcomes (e.g., depression) during the COVID-19 pandemic, compared to those without these histories. However, few studies have examined sleep and most rely on retrospectively-reported pre-pandemic experiences. We tested whether pre-pandemic trauma and stress were prospectively related to worse ecological momentary assessment (EMA)-reported sleep during the pandemic. Methods The sample includes 114 regular drinkers aged 21-30 years from two ongoing studies of alcohol use and sleep who completed a shared assessment battery and a 10-17-day EMA protocol before and during the pandemic (conducted July-November 2020; M=13.9 months after baseline). Participants reported past-month perceived stress (10-item Perceived Stress Scale) and interpersonal traumas (e.g., abuse, conflict), via scores on the “Current Partner” and “Personal” (persons other than spouse/partner) subscales of the Difficult Life Circumstances Scale. The EMA protocol measured daily sleep (total sleep time [TST]; sleep efficiency [SE]), relational stress (1-5 ratings for family, spouse/partner, friends), and alcohol use. Paired t-tests compared pre-pandemic vs. pandemic sleep. Separate linear regressions tested associations between pre-pandemic trauma and stress with average pandemic TST and SE, adjusted for baseline age and sleep, racial identity, assigned sex at birth, time between assessments, and drinking days (averaged across timepoints). Results Participants were on average 23.8 years old (61% female; 7% Asian; 39% Black; 1.8% Mixed race; 0.9% Other race; 0.9% Pacific Islander; 55% White). Average TST increased from baseline to pandemic (7.5 vs. 7.8; t(113)=-2.57, p=.01); no change was observed in SE (95% vs. 94%; t(113)=1.01, p=.31). Pre-pandemic perceived stress (B[SE]=-.003[.001], p=.02) and average EMA-reported family stress (B[SE]=-.04[.02], p=.05) predicted worse pandemic SE. No associations emerged with friend or partner stress, trauma, or TST (ps>.11). Conclusion Pre-pandemic perceived stress (but not trauma nor relational stress) predicted worse sleep during the pandemic. Perceived stress reflects feeling overwhelmed and difficulty coping, which is relevant given dramatic pandemic-related impacts on daily life. The overall accumulation of stress, versus day-to-day stress in specific relationships, may be most detrimental for sleep during the pandemic. Perceived stress is amenable to evidence-based (and remotely-delivered) interventions, including mindfulness-based stress reduction. Support (If Any) R01AA025617 (SLP), R01AA026249 (BPH/SLP), K23HL159293 (KPJ)