Objective To investigate the frailty status of elderly patients with liver cancer after interventional therapy, and to establish and validate a risk assessment model.Methods A retrospective cohort study was conducted, enrolling 200 elderly liver cancer patients from our hospital between January 2023 and June 2025. They were randomly divided into a training set (n = 140) and a validation set (n = 60) at a 7:3 ratio. All patients underwent interventional therapy, and frailty was assessed one month post-treatment using the Tilburg Frailty Indicator (TFI). Least absolute shrinkage and selection operator (LASSO) regression was used to screen the characteristic variables related to frailty in elderly patients with liver cancer after interventional therapy, and the random forest (RF) algorithm was used to establish the risk assessment model. The model performance was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC), calibration curve and clinical decision curve, and was verified using the validation set data.Results Among the 200 patients, the TFI score was 5 (4, 6), and the incidence of frailty was 47.00%. Compared with the non-frailty group, patients in the frailty group were older and had higher rates of complications, as well as higher scores on the age-adjusted Charlson Comorbidity Index (aCCI), Nutritional Risk Screening 2002 (NRS-2002), Cancer Fatigue Scale (CFS), and Pittsburgh Sleep Quality Index (PSQI) (all P < 0.05). The frailty group had more advanced Barcelona Clinic Liver Cancer (BCLC) stages, lower educational attainment, and a lower proportion of married patients than the non-frailty group (all P < 0.05). LASSO regression identified BCLC stage, complications, aCCI score, NRS-2002 score, CFS score, and PSQI score as variables associated with frailty after interventional therapy in elderly patients with liver cancer. Based on these variables, an RF model was developed. The AUC of the RF model were 0.809 (95% CI: 0.687, 0.899) and 0.791 (95% CI: 0.714, 0.855) in the training and validation sets, respectively. Calibration analysis showed Brier scores of 0.187 and 0.176 for the RF model in the training and validation sets, respectively. The calibration curves were generally close to the ideal diagonal line, with slopes approaching 1 and intercepts approaching 0. Decision curve analysis showed that the net benefit of the model was higher than that of the "treat-none" and "treat-all" strategies over threshold probability ranges of 0.13-0.86 in the training set and 0.10-0.90 in the validation set.Conclusion BCLC stage, complications, aCCI score, NRS-2002 score, CFS score, and PSQI score are identified as key risk variables for frailty after interventional therapy in elderly patients with liver cancer. Based on these variables, a reliable risk assessment model was successfully developed using the RF algorithm, which may facilitate the early identification of high-risk patients and provide a reference for optimizing clinical prevention and management strategies.