Radiotherapy is a widely recommended treatment for cervical cancer. The main cause of treatment failure is known to be local recurrence and distant metastasis. This study aimed to develop a predictive model for local recurrence and distance metastasis after radiotherapy, including concurrent chemoradiotherapy and radiotherapy alone, which could play a significant role in treatment plan making. Using clinical characteristics of 1421 cervical cancer patients, among whom all received intensity-modulated radiation therapy, 99.5 % received high-dose-rate brachytherapy and 92.5 % received concurrent sensitization chemotherapy during radiotherapy with an average of 5 cycles, models were generated using machine learning methods of random forest, adaboost and logistic regression. Finally, a combined (stacking) model of the above was generated. The relative importance of factors to outcome was calculated as a percentage contribution to the model. All the models show good discrimination in predicting OS, DFS, LC and DMFS. Out of the four predictive models, the accuracies for all metrics were similar, with the stacking model performing the best (86.0% for OS, 83.7% for DFS, 87.3% for LC and 86.9% for DMFS). Performance was also similar when evaluated under the receiver operating characteristic (ROC) curve (AUCs) (AUC 0.77 for OS, 0.73 for DFS, 0.73 for LC and 0.71 for DMFS). In the final model, the most important variables for OS, DFS and LC were 2018 FIGO stage (14.2%, 17.9% and 16.4%), para-aortic lymph node metastasis (15.6%, 15.8% and 15.8%) and chemotherapy times (13.0%, 16.3% and 22.0%). For DMFS, the most important variables were iliac lymph node metastasis (13.7%) and total EQD2 (13.4%). The machine learning models trained with a handcraft cervical cancer dataset provide excellent performance in quantifying OS, DFS, LC and DMFS and could further help with the clinical prognostic factor’s analysis.