Objective To investigate the risk of postoperative deep vein thrombosis (DVT) of lower extremity in elderly patients with bladder cancer, and determine reasonable early warning measures based on the results of random forest model, so as to provide reference for clinical prevention and treatment.Methods 282 elderly patients with bladder cancer admitted to our hospital from June 2019 to June 2022 were selected and divided into the training group (n=197) and the internal verification group (n=85) according to a ratio of 7∶3.The incidence rate and general clinical data of lower limb DVT in the two groups were analyzed. The training group constructed a lower limb DVT risk warning model based on the results of random forest model, parallel internal and external validation. Results The incidence of lower extremity DVT in 282 elderly patients with bladder cancer was 15.10% (42/278). Among the training group and the internal verification group, there were significant differences in age, BMI, anesthesia time, diabetes mellitus, hyperlipidemia, preoperative FIB, D-D, ALB levels, RAM score and postoperative bed time in DVT-positive and DVT-negative groups (P＜0.05)The first 8 variables of attribute importance score were included in the Random forest model algorithm to establish the lower limb DVT risk warning model. The importance scores of variables were successively preoperative D-D, preoperative FIB, preoperative ALB, preoperative RAM score, age, diabetes mellitus, postoperative bed time, and hyperlipidemia. The prediction efficiency of random Forest model was 0.933. The external verification shows that the predicted results of the model are in good agreement with the actual results.Conclusion The random forest model has a good ability to predict postoperative DVT of lower limb in elderly patients with bladder cancer. Considering age, diabetes mellitus, postoperative bed time, hyperlipidemia, preoperative D-D, FIB, ALB and RAM scores and other factors, it is beneficial to prevent the occurrence of lower limb DVT, which has clinical significance