Abstract: The objective of the present study was to discriminate three kinds of starch by fourier transform infrared spectroscopy(FTIR), deconvolution infrared spectroscopy, second derivative infrared spectroscopy and two-dimensional correlation infrared spectroscopy (2D-IR). The FTIR spectra of starch in the region of 1200- 900 cm- 1 were applied to deconvolution analysis; the peak ratios of 1050/1020 cm- 1 were calculated to analyze the starch ordered structure, the higher the ratios, the more contents of the ordered structure. The infrared spectra in the range of 1800- 700 cm- 1 were selected to perform second derivative infrared spectroscopy analysis. The results showed that the infrared spectra were similar on the whole; the ordered structure from high to low was lotus root starch, potato starch and sweet potato starch; the obvious differences in positions, shapes and absorption intensities of peaks of second derivative infrared spectroscopy were observed in the range of 1607-1487 cm-1, and in the wave number of 1468, 1187, 1055, 931 cm-1. 2D-IR spectroscopy showed that the automatic spectrum peak positions and intensities of three different starch were significantly different in the range of 1150-1500 cm- 1 and 860-1200 cm- 1. The results showed that the macroscopical finger print characters of FTIR and 2D-IR spectra could not only prove the information of main chemical constituents in starch, but also compare the components differences among the similar samples. In conclusion, the multistep IR macro-finger print method were rapid, effective, visual and accurate for starch research.