Objective To compare psychotic major depression(PMD) and non-psychotic major depression(NMD) in clinical features, and treatment methods and effects. Methods 150 cases of severe depression were divided into the PMD group(100 cases) and the NMD group(50 cases). In the PMD group, 41 patients were treated with a single agent, and the other 59 patients were treated with combined agents. The NMD group and PMD mono-therapy group were treated with paroxetine, while the PMD combination therapy group was treated with a combination of oral paroxetine and olanzapine. All groups were scored with the Hamilton Depression Rating Scale (HAMD) at the 1st visit and after 1,3 and 9 months of treatment. Results ① HAMD total scores in NMD and PMD groups at the 1st visit were (53.22±6.00) and (58.30±5.20), respectively(P>0.05). The retardation factor, cognitive factor, and suicide factor in the PMD group were higher than those in the NMD group. ②After 9 months of treatment, the HAMD score in the PMD monotherapy group was higher than that in the PMD combination therapy group. The score in the PMD mono-therapy group was higher than in the NMD group. ③At the 1st test with the Pittsburgh Sleep Quality Index(PSQI), NMD and PMD had no significant difference in sleep disorders(P<0.05), while they were statistically different from the healthy controls. After 9 months of treatment, NMD and PMD were statistically improved in total scores and all the main factor scores (P<0.05), and they were still statistically different in first factors “subjective sleep quality” and “daytime dysfunction”, and the secondary factor “wake up at night or wake too early”. Conclusions ①PMD is as serious as NMD. Retardation, cognitive impairment and suicidal thoughts in PMD patients are more obvious than those with NMD,while depression in NMD patients is more obvious.②For PMD, treatment with paroxetine and olanzapine is more effective than paroxetine alone. Compared with PMD patients, paroxetine monotherapy is more suitable for NMD patients.③Major depression has a decline in comprehensive sleep function. After effective treatment, sleep function can be significantly improved, but part of sleep dysfunction still remains. Mental symptoms cannot affect sleep function damage and recovery.