〔Abstract〕 To explore the risk of endometrial polyp (EP) based on lipid metabolism and vaginal micro- ecology combined with uterine volume line drawing model. Methods 143 EP patients treated by hysteroscopic sur- gery were selected as the experimental group , and 113 healthy women were selected as the control group at the same time. The data were randomly divided into training set and validation set according to the ratio of 7 : 3. The clinical data of the two groups were collected and recorded , and t/χ2 test , LASSO regression and multifactorial lo- gistic regression analysis were used to screen the independent risk factors , construct the prediction model , and draw the column line graph. The performance of the model was evaluated by applying subject operating characteristic (ROC) curves , calibration curves , Hosmer-Lemeshow test and clinical decision-making (DCA) curves. Results Multifactorial logistic regression analysis showed that total cholesterol ( TC) , low-density lipoprotein cholesterol (LDL-C) , vaginal microecological balance , and uterine volume were independent risk factors for the development of EP. ROC curve analysis showed that the AUC values of the training and validation sets of the column line graph model were 0. 935 and 0. 887 , respectively , and its sensitivity and specificity were 90. 21% , 83. 46% and 86. 29% , 80. 66% respectively , The Hosmer-Lemeshow test showed that the model fits well ( training set : χ2 = 2. 261 , P = 0. 840 ; validation set : χ2 = 4. 837 , P = 0. 441) and the calibration curves of the training and validation sets were close to the ideal curves , which indicated that the model had good prediction accuracy; the analysis of DCA curves of the training and validation sets both showed that the column-line graph model had a good clinical benefit rate in predicting EP. Conclusion TC , LDL-C , vaginal microecological balance and uterine volume are independent risk factors for EP , and the column-line diagram model constructed by the model has high clinical ben- efit , calibration and accuracy in predicting the risk of EP.