Medicinal gelatin hollow hard capsule has been widely used in people's daily life, but the toxic capsule event has drew public concern on its safety issue in 2012, since the casings were made with an industrial gelatin with high amounts of metal chromium. In this paper, the traditional method, atomic absorption spectrophotometer, was used for detecting medicinal gelation hollow hard capsule as a control group; then hyperspectral image was used for detecting 347 normal capsule samples and 701 toxic capsule samples, which were analyzed and discussed from different image bands and regions of interest (ROIs). At the beginning, hyperspectral processing was used for dimension reduction by Principal Component Analysis (PCA) method, and then two kinds of capsules spectral data were obtained to do qualitative analysis according to ROIs. Partial Least Squares Discriminant Analysis(PLS-DA) and Principal Component Analysis- Artificial Neural Network(PCA-ANN) were used to identify Chromium-laced capsule based on spectral data processing, of which 60% sample are used for modeling, the remaining 40% are used for predicting. In the end, we can draw a conclusion that if chosen 4 Latent Variables (LVs) as input features in the PLSDA model, high classification accuracy would be achieved, with the accuracy of 100%. The correlation coefficient of cross validation and sample prediction are 0.923 and 0.972 respectively; sensitivity and specificity are 100% as well. The result prove that it is feasible to use hyperspectral data for chromium-laced capsule qualitative analysis, the method greatly reduces the complexity of the traditional detection and try to provide a new thought in chromium detection of capsule. © 2016, Asian Association for Agricultural Engineering. All rights reserved.