Abstract Introduction Sleep architecture is altered in insomnia, including reduced rapid eye movement (REM) sleep. As REM sleep supports cognition and emotional regulation, its preservation has potentially important clinical implications. Treatment with lemborexant (LEM), a dual orexin-receptor antagonist, led to increased REM sleep compared with placebo (PBO) in adults with insomnia. This investigation evaluated the comparative effects of LEM on REM sleep parameters between Asian and non-Asian adults with insomnia, thereby evaluating potential effects of race. Methods This analysis incorporated data from 3 independent, 1-month, randomized, double-blind, PBO-controlled, parallel-group studies. E2006-J086-311 (Study 311; NCT04549168) included Chinese participants ≥18y. E2006-J082-204 (Study 204; NCT05594589) included Korean participants 19–80y. E2006-G000-304 (Study 304; NCT02783729) enrolled participants ≥55y (females)/≥65y (males) of any race; only non-Asian participants are presented here. Polysomnographic assessment of REM sleep and REM latency (REML) was conducted at baseline (during single-blind PBO run-in) and following 1 month of treatment with LEM 10mg (LEM10) or PBO. Results Mean (SD) baseline REM for LEM10 and PBO was 68.65(19.167) and 69.85(17.091) min, 67.37(28.695) and 58.96(36.077) min, and 61.58(20.409) and 65.32(19.591) min for Study 311 (LEM10, n=93; PBO, n=100), Study 204 (LEM10, n=26; PBO, n=13), and Study 304 (LEM10, n=264; PBO, n=206), respectively. At 1 month, least squares mean (LSM) (SE) CFB in REM for LEM10 and PBO was 24.05(3.451) and 8.59(3.176) min (P< 0.0001), 28.06(5.325) and 12.77(7.529) min (P=0.1035), and 21.82(1.368) and 5.10(1.528) min (P< 0.0001) for Study 311, Study 204, and Study 304, respectively. Mean (SD) baseline REML for LEM10 and PBO was 98.26(41.392) and 95.59(33.729) min, 125.63(68.727) and 131.42(91.967) min, and 100.14(54.320) and 99.60(51.803) min for Study 311, Study 204, and Study 304, respectively. Mean (SD) CFB in REML for LEM10 and PBO was −38.47(45.677) and −7.20(37.393) min (P< 0.0001), −42.15(73.471) and −0.83(59.514) min (P=0.0709), and −38.16(56.103) and −7.28(62.479) min (P< 0.0001) for Study 311, Study 204, and Study 304, respectively. Conclusion The magnitude of changes in REM sleep and REML was comparable between Asian and non-Asian participants, demonstrating LEM’s consistent therapeutic effect across diverse patient populations. Given REM sleep’s established role in cognition and emotional regulation, these findings suggest potentially significant clinical implications warranting further evaluation. Support (if any) Eisai