Purpose <p>Athletes undergo intense training to enhance performance. However, such exertion can lead to overtraining syndrome (OTS). OTS can lead to a decline in performance, muscle pain, and prone to muscle injuries. In-situ monitoring of muscle for fatigue helps in the prevention of OTS and ensures safe training.</p> Objective <p>This study presents a machine learning framework for real-time monitoring of muscles for athletes during strengthening exercises using surface electromyography (sEMG).</p> Methods <p>sEMG signals were recorded from the hamstring muscles of collegiate athletes. sEMG features reflecting amplitude and spectral characteristics, along with load intensity information, were used to train support vector machine (SVM) classifiers with different kernels. These models were also validated on unseen data to evaluate their performance and real-world applicability.</p> Results <p>All trained SVM models achieved an accuracy above 90% in the prediction of muscle conditioning. Furthermore, when evaluated on unseen test data, the linear and radial basis function SVMs achieved superior performance, with accuracies around 96%, demonstrating the robustness of the proposed approach and its potential applicability for real-time fatigue detection.</p> Conclusion <p>These findings of the proposed pipeline enable framework for continuous real-time fatigue monitoring and can assist coaches in tracking muscle status, optimizing training, and preventing overuse injuries.</p>