Fritillariae thunbergii Bulbus, as one of the famous “Zhe Ba Wei”, is often used as expectorant and antitussive medicine. Sulfur fumigation is commonly good for whitening, mothproofing and extending the shelf-life of medicines. However, excessive sulfur fumigation will affect the quality of traditional Chinese medicines, even damage human health. Therefore, using non-destructive detection method of Fritillariae thunbergii Bulbus with sulfur fumigation is beneficial to monitor the quality of traditional Chinese medicines. This paper used near infrared spectroscopy (NIR) (900～1 700 nm) combined with chemometrics to detect the Fritillariae thunbergii Bulbus fumigated with different concentrations of sulphur. “Boxplot” was adopted to analyze the difference among the samples in the range of 1 000～1 100 nm. Principal component analysis (PCA) was also used to cluster analysis of six different samples. Then successive projection algorithm (SPA) was applied to extract ten different characteristic wavelengths and establish the partial least squares-discriminant analysis (PLS-DA) models. Results showed that PLS-DA model had the similar results with full spectra analysis. Furthermore, NIR combined with chemometrics can be used to analyze the Fritillariae thunbergii Bulbus fumigated with different concentrations of sulphur. It provided a theoretical reference for the analysis of the quality of Fritillariae thunbergii Bulbus and the design of the portable instruments.