Attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy shows great potential for biomedical applications, especially in clinical hematology, due to its rapid, accurate, and non-invasive nature. The technique captures molecular vibrational spectra, yielding detailed chemical information of nucleic acids, proteins, and lipids in biological samples critical for disease screening and diagnosis. Extensive research demonstrates the efficacy of ATR-FTIR in diagnosing diseases, such as thalassemia, AIDS, breast cancer, ovarian cancer, and brain tumors. Combining ATR-FTIR with chemometrics enables accurate disease screening and differential diagnosis. ATR-FTIR spectroscopy combined with partial least squares regression model was used to quantitatively analyze thalassaemia screening indices in human peripheral blood samples，such as mean erythrocyte haemoglobin content, mean erythrocyte volume and haemoglobin, and the sensitivity and specificity of screening reached 100.0% and 95.3%, respectively. ATR-FTIR spectroscopy combined with linear discriminant analysis method based on genetic algorithm were used to identify the characteristic peaks at 1,558 cm^-1 (amide Ⅰ band), 1,506 cm^-1 (cyclic group) and 901 cm^-1 (phosphodiester stretching band) in the blood samples of pregnant women, the sensitivity and specificity for diagnoding human immunodeficiency virus (HIV) infection was 89% and and 92%, respectively. In breast cancer, ATR-FTIR coupled with principal component regression （PCR） methods reached sensitivity and specificity of 92.3% and 87.1%, respectively. Furthermore, ATR-FTIR spectroscopy is applicable in other biomedical domains, such as detecting cellular and histological samples and classifying disease severity. Despite its promise, challenges like environmental interference and sample contamination persist. Future advancements and optimizations in ATR-FTIR spectroscopy are anticipated to enhance its role in clinical hematology and extend its applicability to a broader spectrum of diseases.