Abstract Introduction Cognitive dysfunction in individuals with schizophrenia (SZ) represents a major determinant of poor functioning and may arise in part due to deficits in sleep quality that have been reported in in this population. In the current project, we investigate the effect of acute sleep restriction on a sleep-dependent emotional memory task. Methods Six individuals with medication-controlled SZ encoded consensus-rated emotionally valent and neutral pictures from the International Affective Picture System in the evening before 2 counterbalanced, polysomnographically-recorded nights containing either ad libitum sleep (~11pm to 7am) or sleep restriction (SR) to 4 hours between 3 and 7 am. Image recall occurred within 1 hour of wake time. We compared memory performance for negative emotional/neutral memory and PSG measures across the two sleep conditions with paired signed-rank tests. Results Mean age was 38 years and 50% were women. Acute sleep restriction resulted in reduced total sleep time (Adlib: median 372 min. (interquartile-range 53) vs SR: 209 (16) min., p=0.03), including reduction in duration of N1 (Ad lib: 44 min. (44) vs. SR: 29 min. (33), p=0.03), N2 (Ad lib: 158 min. (86) vs. SR: 60 min. (33), p=0.03) and REM sleep (Ad lib: 18 min. (5) vs. SR: 13 min. (13), p=0.03). We did not observe differences in N3 duration, sleep efficiency, and arousal index. Measures of breathing such as AHI4%, AHI3A, and SpO2 nadir were not significantly different between conditions. Compared to SR, the false alarm rate for emotional images was increased in ad lib sleep (n=6, p=0.03) with no difference in false alarm rate for neutral images (p=0.19). Correct rejection rate for emotional images was increased in SR compared to ad lib sleep (p=0.03) with no difference in neutral images (p=0.19). No other change was observed in accuracy of recall for emotional or neutral, miss rate for emotional or neutral and d-prime for emotional or neutral images. Conclusion In subjects with SZ, unrestricted sleep was associated with a greater false alarm rate for emotional, but not neutral scenes, suggesting a deficit in veridical memory or generalization of emotionally encoded information. Support (if any) R21MH126357, R01AG066870, R01AG056682, K24HL109156, AASMF Focus award/AARG-D