Introduction The mechanism underlying the development of excessive daytime sleepiness (EDS) in obstructive sleep apnea (OSA) remains unclear. Thus, we characterised the association between sleep architecture and EDS in individuals with and without OSA. Methods 1876 suspected OSA patients reporting daytime sleepiness underwent in-lab polysomnography and next-day multiple sleep latency test (MSLT). We investigated sleep architecture characteristics in OSA (apnea-hypopnea index [AHI]≥5, n=1508) and non-OSA (AHI<5, n=368) patients with and without EDS (MSLT≤10-minutes and MSLT>10-minutes). Sleep architecture was quantified by N1, N2, N3, REM, total sleep time (TST), and wake after sleep onset (WASO). Results OSA patients with EDS had less N1 (median: 10.0 vs 12.0-minutes, p<0.05) and N3 (60.2 vs 71.5-minutes, p<0.05) sleep, less WASO (43.0 vs 51.0-minutes, p<0.05), more N2 (210.5 vs 189.0-minutes, p<0.05) sleep, and higher TST (365.0 vs 351.5-minutes, p<0.05) than non-EDS patients. No difference was observed in the amount of REM (70.0 vs 67.5-minutes, p=0.46) sleep. Non-OSA patients with EDS had less WASO (31.5 vs 40.0-minutes, p<0.05), more N2 (188.5 vs 177.5-minutes, p<0.05) sleep, and higher TST (374.5 vs 359.0-minutes, p<0.05) and the amount of REM sleep (76.5 vs 68.5-minutes, p<0.05) than non-EDS patients. Discussion While EDS was associated with less WASO and greater TST and N2 in both OSA and non-OSA groups, there was also characteristic differences. Non-OSA patients with EDS have greater REM sleep duration, while OSA patients with EDS have less N3 sleep. There is likely a complex bi-directional relationship between sleep architecture and EDS reflecting acute/chronic sleep disruption and compensatory mechanisms.