Each year, numerous motor vehicle drivers (MVD) suffer injuries or fatalities in traffic accidents. Many factors are related to the severity of injuries sustained by MVD in traffic accidents. The primary objective of this study is to investigate the interaction of factors influencing injury severity of MVD and to predict injury severity. Using the traffic accident data collected from a city in China between 2018 and 2023, we employed the multinomial logistic regression model to identify significant factors affecting injury severity of MVD. Subsequently, these significant factors were incorporated into a stepwise logistic regression model to analyze the relationship between the interaction of factors and injury severity of MVD. Furthermore, the significant factors in the logistic regression model were taken as independent variables, and Random Forest, XGBoost, and LightGBM were used to construct the prediction model of injury severity of MVD. The results indicate that 13 factors have significant meaning for MVD. The stepwise logistic regression model reveals that nine interactions have the most significant impact on injury degree of MVD. The quantitative analysis of the relationship between injury severity and interaction indicates individuals aged 66 and above with abnormal vehicle status, individuals aged 26–45 with guardrails, individuals aged 46–65 and special road, vehicles with normal status and no lighting at night, vehicles with normal status and other driving state, green plants and green belts and early morning, sand, gravel, or dirt roads and cloudy day, mountain and no lighting at night can lead to increased injury severity of MVD. Among the prediction models, the XGBoost model with SMOTE sampling achieved the highest accuracy, macro-averaged precision, macro-averaged recall, and macro-averaged F1 score, with an accuracy of 0.907. The findings of this study provide valuable insights for traffic management authorities to improve traffic safety and reduce the severity of MVD injuries in traffic accidents.