Infectious diseases such as acute nasopharyngitis, acute pharyngitis, and acute tonsillitis remain major public health issues, especially in primary healthcare facilities with limited resources like Puskesmas Gunungsari. This study aims to develop a machine learning-based classification model to detect infectious diseases using patient medical data. The evaluated models include Random Forest, Decision Tree, Support Vector Machine (SVM), and Neural Network, with performance assessed using k-fold cross-validation ranging from 5 to 10 folds. Evaluation results show that the Decision Tree consistently achieved the best performance, with an accuracy of approximately 91.7% to 91.9% and an F1-score ranging from 91.9% to 92.3% on cross-validation data, as well as a test accuracy of 94.7% and an F1-score of 95.0%. The Random Forest model also demonstrated good and stable performance, with accuracy between 90.5% and 90.7%. Meanwhile, SVM and Neural Network produced lower results, with maximum accuracy of around 77.0% and 71.7%, respectively. Overall, the findings demonstrate that the Decision Tree model is the most effective for supporting early diagnosis of infectious diseases at Puskesmas Gunungsari, providing superior classification capabilities compared to other models.