Opisthorchis viverrini (Ov) infection is a major public health concern in the Greater Mekong Subregion, with cats as key reservoir hosts. Although machine learning (ML) is widely used in human medicine, its application in veterinary epidemiology remains limited. This study aimed to develop interpretable ML models to predict Ov infection and to identify key risk factors in cats using data from 175 households in endemic areas. Five ML algorithms-Classification Tree, Random Forest, Ridge Logistic Regression (Ridge LR), eXtreme Gradient Boosting, and Support Vector Machine (SVM)-were optimized using feature selection methods, hyperparameter tuning, and SMOTE. The results demonstrated that Ridge LR with Minimum Redundancy Maximum Relevance (mRMR), tuned Ridge LR, and tuned SVM achieved reliable performance, with ROC-AUC values exceeding 0.7. Specifically, default Ridge LR with mRMR achieved strong balanced accuracy (0.722), while tuned Ridge LR attained the highest sensitivity (0.667). Tuned SVM with mRMR yielded a test ROC-AUC of 0.723 and PR-AUC of 0.473, along with a balanced accuracy of 0.682. SHapley Additive exPlanations (SHAP) analysis identified key risk factors, including residence in flooded areas, feeding fish scraps to cats, and annual rainfall, emphasizing the role of environmental factors in Ov transmission. These findings highlight the potential of ML in veterinary epidemiology and emphasize the importance of selecting appropriate methods based on data characteristics. The study suggests that targeted, risk-based interventions focusing on these key risk factors are crucial for effective Ov control in endemic regions.