Osteoporosis is a major health issue which can lead to serious clinical conditions including fractures. The resulted fracture can be a major factor of disability or even mortality for the elderly. Therefore, an early diagnosis of osteoporosis can help in detecting and preventing such fractures. Prevention of osteoporosis can significantly reduce economic and health costs. The purpose of this paper is to develop machine learning models with feature selection techniques to predict osteoporosis based on the provided features. A total of 1958 patients were included with features related to demographic characteristics, lifestyle, and medical indicators. Seven different classifiers with 20 distinctive feature selection techniques were deployed. Out of the deployed algorithms, SVM classifier with fsSubsetEval feature selection and SVM, KNN and RF classifiers associated with the wrapper method using NB as base classifier demonstrated superior performance, which achieved an accuracy of 91.4%. The performance of predictive algorithms was further validated and analyzed. The results showed that our predictive model has an AUC of 0.925, indicating good predictive performance. The choice of predictive algorithms in this study was essential in ensuring robust prediction performance. The predictive model is critical for identifying individuals at risk of osteoporosis, enabling early intervention and prevention strategies.