Near-infrared (NIR) Surface Enhancement Raman scattering (SERS) is a highly sensitive and selective surface-matter detection technique that uses oxygenated hemoglobin in human blood to detect the corresponding NIR-SERS spectra. The spectral data were detected and analyzed using a combined multivariate statistical method. Firstly, the NIR-SERS spectra of oxygenated hemoglobin in 30 healthy subjects and 30 patients with nasopharyngeal carcinoma were screened, and it was found that the average NIR-SERS spectra of oxygenated hemoglobin in NPC patients were at the positions of 340, 479, 817, 1 127, 1 215, 1 347, 1 425, 1 587 cm-1. The area and intensity of the peaks were relatively small, especially at the positions of 479, 817 and 1 587 cm-1. In addition, when the NIR-SERS peaks of healthy individuals were at 335, 471, 1 338 and 1 429 cm-1, the red shift to 340, 479, 1 347 cm-1and blue shift to 1425 were observed in NPC patients.The multivariate statistical model of PCA-T (principal component analysis combined with independent variable t-test) was established to obtain the three most significant PCA scores: PC1, PC3, and PC7, two-dimensional and three-dimensional scatter plots were made for the three PCA scores, and the healthy people and NPC patients were divided into different regions. A working characteristic curve (ROC) was used to evaluate the reliability of the PCA-T analysis method and verify its reliability. The results showed that the vibrational changes of oxyhemoglobin molecules, such as pyrrole ring and vinyl group, were abnormal, and the changes of L-Phenylalanine were also significant. The nasopharynx of the human body likely has pathological changes, resulting in the patients bodys more vigorous metabolism of amino acids, increased sugar consumption, and other causes. In conclusion, the combination of near-infraredsurface-enhanced Tunku Abdul Rahman scattering spectroscopy with multivariate statistical analysis has the potential to accurately, effectively, and rapidly differentiate healthy human samples from those of nasopharyngeal carcinoma, this study provides a preliminary experimental basis for the application of NIR-SERS spectroscopy in nasopharyngeal carcinoma and other medical studies and may be a potentially sensitive and reliable tool for early clinical cancer diagnosis.