Abstarct: Rapid, accurate, and non-invasive diagnosis of diseases is one of the important needs of modern medicine. In recent years, the analysis of human breath, due to containing volatile organic compounds, has attracted attention as a promising method for identifying biomarkers associated with diseases such as cancer, diabetes, respiratory diseases, and mextabolic disorders. However, the low concentration of these compounds in breath, the presence of high humidity, the interference of various gases, and the need for high selectivity have faced the use of conventional sensors with limitations. In this review study, nanomaterial-baxsed breath gas sensors have been investigated in terms of the type of nanomaterial, mechanism of operation, performance indicators, and medical applications. Nanomaterials such as mextallic nanoparticles, mextal oxides, graphene, carbon nanotubes, and nanocomposites, due to their high specific surface area, suitable conductivity, surface modification capability, and rapid response, play an important role in increasing the sensitivity and selectivity of breath sensors. Also, in this study, the mechanisms of detecting respiratory biomarkers, the effect of real breath conditions, the performance of sensor arrays and electronic nose, and the role of data analysis methods and machine learning in improving diagnostic accuracy have been investigated. The results of this review show that nanomaterial-baxsed breath sensors can be used as low-cost, portable, and non-invasive tools for early diagnosis and disease monitoring. Nevertheless, challenges such as standardization of sampling, sensor stability, the effect of humidity, interference of gaseous compounds, and the need for extensive clinical validation still prevent the full entry of this technology into clinical applications. Therefore, the development of more stable and more selective sensors and their integration with intelligent data analysis systems can shape the future path of this field.