Objective To evaluate the accuracy of the Self-Rating Depression Scale and Beck Depression Inventory in diagnosing depression and distinguishing the severity of depression. Methods From August 2018 to August 2020， a total of 200 voluntary patients with mood and sleep problems in Chongqing Mental Health Center were enrolled. The diagnosis of depression and the severity of depression were assessed according to the ICD-10 entry criteria， and the Self-Rating Depression Scale and Beck Depression Inventory were evaluated at the same time. The accuracy of the two depression scales was evaluated by area under ROC curve. Results There was a significant correlation between the Self-Rating Depression Scale and Beck Depression Inventory （r=0.848， P ＜ 0.001）. The area under the ROC curve for diagnosing depression and distinguishing the severity of depression were similar and there was no statistical difference between the two scales （P ＞ 0.1）. With the aggravation of depression， the area under ROC curve gradually increased （0.754/0.761-0.904/0.911）， and Youden index gradually increased （0.498/0.403-0.768/0.751）. The positive likelihood ratio increased gradually （2.71/2.14-9.63/10.1）， and the positive predictive value increased gradually but was not ideal （17.2%/17.8%）. Conclusions The accuracy of the Self-Rating Depression Scale and the Beck Depression Inventory in diagnosing depression and distinguishing the severity of depression is similar， and the more severe the depression is， the more accurate the evaluation of the two scales will be. The two scales have specific purposes and specific populations， so they cannot be used arbitrarily.