In this study we have determined whether Raman and infrared spectroscopy of blood plasma differentiates Alzheimer's disease (AD) from normal aging of healthy controls. Spectroscopic analysis was conducted on blood plasma samples from 8 mild AD, 16 moderate AD, 11 severe AD, and 12 normal elderly control persons using Fourier transform spectrometers and a near-infrared laser beam as excitation source for Raman spectroscopy. Spectra were processed employing discriminant analysis to determine whether band areas and frequency-intensity relationships might reveal biochemical differences associated with AD. Seven spectral biomarkers were identified in the Raman regions of 1700-1600 cm-1 (protein secondary structure), 980-910 cm-1 (protein α-helices), 790-730 cm-1 (protein tertiary structure), and 440-390 cm-1 (protein backbone) and in the infrared regions of 1700-1600 cm-1 (protein secondary structure) and 1150-1000 cm-1 (oxidative stress). This discriminant analysis model differentiated AD from normal aging of elderly control persons with a sensitivity of 89% and specificity of 92%. Moreover, specificity increases to 100% for the detection of mild AD. In summary, our results open the possibility of using this spectroscopic approach as a non-invasive, rapid, and relatively inexpensive procedure for early accurate diagnosis of AD.