Objective To compare the differences of immune and stress related factors between bipolar depression and unipolar depression， and to further explore potential peripheral blood biomarkers of these two kinds of depressions. Methods The inpatients of Beijing Anding Hospital Affiliated to Capital Medical University from January to December 2018 were selected as the research objects. 933 patients with monophasic depression and 546 patients with bipolar depression were included in this study. Demographic data and biochemical indicators were compared between the two groups. The biochemical indicators included C-reactive protein （CRP）， complement 3 （C3）， complement 4 （C4）， immunoglobulin M （IgM）， immunoglobulin G （IgG）， immunoglobulin A （IgA）， adrenocorticotropic hormone （ACTH） and cortisol. Logistic regression analysis was used to control the confounding factors and clarify the correlation between the two types of depression and various biological indexes. Receiver operating characteristic （ROC） curve was used to analyze the predictive value of the above demographic and biological indexes related to bipolar depression. Results Bipolar depression and unipolar depression were different in the terms of C3 （Z=3.41， P=0.001）， IgG （Z=-2.04， P=0.041）， IgA （Z= -4.30， P＜ 0.001） and ACTH （Z=2.03， P=0.042）， with statistical significance （P＜ 0.05）. Logistic regression analysis showed that age （OR=0.98， 95%CI=0.97-0.98）， gender （OR=0.62， 95%CI=0.49-0.78）， C3 （OR=3.45， 95%CI=1.90-6.27） and IgA （OR=0.80， 95%CI=0.71-0.91） were associated with bipolar depression. ROC analysis showed that C3， IgA and age alone had poor accuracy in the classification of bipolar depression and monophasic depression， and the area under the curve （AUC） was 0.55-0.63. After logistic regression modeling， the AUC of predicting bipolar depression was 0.66 （P＜ 0.05）， the sensitivity was 65%， and the specificity was 63%. Conclusions Younger age， higher levels of C3， and lower levels of IgA are related factors of bipolar depression. Compared with a single biological index， the combined diagnostic model including C3， IgA， gender and age is more effective in distinguishing monophasic depression from bipolar depression.