Objective To explore the relationship of sleep disturbance and depression in patients with stroke. Methods Totally 138 convalescent patients with primary stroke from January 2016 to December 2017 were enrolled the study. Hamilton Depression Scale (SDS) and Pittsburgh sleep quality index (PSQI) were performed to evaluate the depression and sleep quality, and post-stroke depression (PSD) was diagnosed by the SDS score. The relationship between sleep disturbance and PSD was analyzed. Multivariate logistic regression analysis was performed to investigate the related factors of PSD. Pearson correlation analysis was used to investigate the relationship between the score of SDS and PSQI. Results PSD incidence rate of patients divorced or widowed and with a history of diabetes, hemorrhagic stroke, stroke in the subcortical, left hemisphere, no regular rehabilitation exercise, significant self-care disability and sleep disorders is increased (P < 0.05). Multivariate logistic regression results showed that marital status [Ol ^ R=2.517, (95 CI: 1.133, 5.594), P =0.023], types of stroke [Ol ^ R=1.126, (95 CI: 1.005, 1.262), P =0.041], rehabilitation exercise status[Ol ^ R=1.283, (95 CI: 1.024, 1.607), P =0.030], significant selfcare disability [Ol ^ R=2.942, (95 CI: 1.291, 6.707), P =0.010] and sleep disorders [Ol ^ R=1.585, (95 CI: 1.105, 2.273),P =0.012] were the key independent risk factors for PSD (P < 0.05). The SDS scores of all patients were positively correlated with the scores of PSQI subjective sleep disorder, sleep efficiency, sleep disturbance, and daytime dysfunction (P < 0.05). Conclusions Sleep disorders are prone to occur in stroke patients during rehabilitation period. Sleep disorders are correlated with PSD. The worse the quality of sleep, the worse the depression.