Based on the differences between auto-fluorescence spectra of plaque and dental tissue, a quantitative analysis method of plaque using auto-fluorescence imaging was presented. Auto-fluorescence spectrum of tooth samples excited by 405 nm was measured, on the basis of analyzing the correlation between color parameters of fluorescence and dental plaque content, R/B of fluorescence image was proposed as quantitative assessment parameter of dental plaque content, meanwhile long-wavelength pass filter with central wavelength of 490 nm was used to filter fluorescence spectral bands. Finally, experimental setup was built to obtain the auto-fluorescence image. Quantitative analysis results of three different levels of plaque content and distribution was obtained by using R/B value as segmentation thresholds, which confirmed with visual examination results.