As a novel and effective spectroscopic analytical tool, Fourier transform infrared spectroscopic imaging (FTIRI) can obtain composition and structure information of samples at high resolution and has been applied in biomedical photonics field. FTIRI is employed to analyze the characteristic IR bands of principal components of five articular cartilages and their attribution by integrated and second derivative methods. The depth dependent profiles of integrated absorbance and second derivative spectral intensity of these IR bands are lineally fitted to those of the concentrations of collagen and proteoglycan that are predicted by principal component regression algorithms.The correlation coefficients by linear fit suggest that the correlation coefficient (R) between the integrated absorbance of amide Ⅰ (amideⅡ) and collagen concentration (R A-amide Ⅰ& RA-amide Ⅱ) is much higher than that between amide Ⅲ absorbance and collagen concentration (RA-amide Ⅲ), namely, RA-amide Ⅰ, RA-amide Ⅱ ＞R A-amide Ⅲ, as well as the correlation coefficient between the secondderivative spectral intensity of sugar band and PG concentration better than that between integrated absorbance of sugar band and PG concentration, RD-sugar ＞RA-sugar. It is concluded that the integrated intensities of amide I and amide II are very suitable to qualitatively represent the content and depth dependence of collagen. The second derivative spectral intensities of sugar band is much fit to represent the contents and depth dependences of PG, respectively. The integrated or the second derivative spectral intensity of amide Ⅲ band is unfit to represent either collagen or PG content in cartilage. This study provides a simpler and quicker method for the investigation on cartilage and early-stage osteoarthritis, which will be helpful for the diagnosis and recovery monitoring of early-stage osteoarthritis.