This study evaluates and compares the performance of various machine learning models in detecting anomalies in unmanned aerial vehicles (UAVs). A simulated UAV fault dataset is utilized, and the models are trained and tested using a 70:30 train-test split. Among the evaluated models, Random Forest demonstrates the highest classification accuracy, followed closely by Decision Tree (DT), while Support Vector Machine (SVM) and Logistic Regression (LR) exhibit comparatively lower performance. However, despite its superior accuracy, Random Forest incurs a longer execution time than the DT model. These findings suggest that while Random Forest is a robust choice for accuracy-focused applications, practical deployment should also consider computational efficiency. Balancing accuracy and execution time remains crucial when selecting models for real-time UAV monitoring systems.