Objective To build a predictive model of depression with secondary mild cognitive impairment (MCI) in elderly patients based on current clinical diagnosis and treatment technology,and to analyze its application.Methods Elderly patients with depression hospitalized in three hospitals were consecutively included in our study from September 2013 to December 2015 for collecting relevant clinical data,and followed up for 18 months to confirm a prognosis.The follow-up results were used to predict influencing indices for secondary MCI risk,and to verify judgement effectiveness of the critical value of the relevant indices on the window of time of the secondary MCI.Results A total of 216 elderly patients with depression were included in this study,of whom 9 patients were lost to follow-up.Finally,27 patients had secondary MCI,and 180 patients had normal cognitive function during the follow-up period.Cox multiple regression analysis showed that the risk model of secondary MCI in elderly patients with depression was composed of age (HR:1.30,95 ％ CI:1.12-1.64,P =0.03),education years (HR:0.56,95 ％ CI:0.41-0.80,P =0.01),regular psychological treatment (HR:0.73,95％ CI:0.58-0.92,P=0.03),and BSSI scale (HR:1.24,95％ CI:1.08-1.56,P=0.03).Age and BSSI scale were risk factors,while education years and regular group psychotherapy were protective factors.For an elderly patient with depression who was characterized by age ≥ 72.3 years,education years ＜8.3 years,and BSSI scale ≥75.1,the window of time for secondary MCI was shorter,and these critical values of the independent factors had significant judgement effectiveness.Conclusions Age,education years,regular psychological treatment,and BSSI scale are independently influencing factors for secondary MCI in elderly patients receiving the treatment for depression.Furthermore,age ≥72.3 years,the education period ＜8.3 years,and BSSI scale ≥75.1 points are critical values of secondary MCI.