Background: The consistency of effects of lemborexant (LEM), a dual orexin-receptor antagonist, on sleep maintenance variables across 2 phase 3 studies with contrasting populations was compared. Methods: E2006-G000-304 (Study 304; NCT02783729) and E2006-J086-311 (Study 311; NCT04549168) were 1-month, randomized, double-blind, placebo (PBO)-controlled studies evaluating LEM 10mg (LEM10) in adults with insomnia disorder. Global Study 304 (N=1006; PBO, n=208; LEM10, n=269) enrolled participants of any race (≥55y); Study 311 (N=193; PBO, n=100; LEM10, n=93) participants were exclusively Chinese (≥18y). Pairs of polysomnograms were conducted at baseline and after the first/last 2 doses of the 1-month treatment. Change from baseline in sleep efficiency (SE [%]), wake-after-sleep-onset (WASO [min]), and total-sleep-time (TST [min]) were analyzed. Results: Mean baseline sleep parameters: Study 304: SE, 67.9–68.9; WASO, 111.8–114.8; TST, 325.1–330.7; Study 311: SE, 69.4–70.3; WASO, 79.3-–85.8; TST, 333.2–336.7. Least squares mean [standard error] increases from baseline were significantly larger with LEM10 vs PBO (P<0.001) for SE (Study 304, 8.0 [0.7]; Study 311, 7.1 [1.4]) and TST (38.9 [3.7]; 32.8 [6.9]), as were decreases in WASO (-25.4 [3.1]; -17.8 [4.8]). Most treatment-emergent adverse events were mild–moderate. Conclusions: Short-term LEM10 treatment consistently improved objective sleep maintenance in patients with insomnia of different races.