Background: Many studies have indicated associations between impaired nasal breathing and sleep disorders. However, the precise nature of the relationship between nasal patency and sleep remains unclear. Purpose: We analysed the effects of nasal patency on sleep architecture and breath in nasal obstruction-predominant obstructive sleep apnoea (NO-OSA) patients by applying nasal decongestant. Material and methods: A randomized, placebo-controlled double-blind crossover study was performed in OSA patients with chronic nasal obstruction and without obvious pharyngeal narrowing. All OSA patients (confirmed by polysomnography) were recruited and completed 2 overnight studies (randomly applying oxymetazoline or placebo). Data collected after oxymetazoline or placebo treatments were compared. The ClinicalTrials.gov identifier is NCT03506178. Results: Compared with placebo, oxymetazoline resulted in significant increase in rapid eye movement sleep (p = 0.027) and reduction of stage 1 sleep (p = 0.004), as well as arousal index (p = 0.002). Moreover, great improvements in apnoea/hypopnea index (AHI) were observed (p < 0.001); AHI in the supine position was significantly reduced (p = 0.001). Oxygen saturation during sleep was increased significantly [mean oxygen saturation (p = 0.005) and lowest oxygen saturation (p = 0.024)]. Oxygen desaturation index was significantly reduced (p < 0.001). Conclusions: Improving nasal patency by decongestant could improve sleep quality, AHI, and oxygen saturation level during sleep.