A rapid, accurate and no sample preparation required method was developed by near infared spectroscopy(NIR) to identify 41 (seven categories) solid excipients widly used in pharmaceuatical industry. 6 batches samples in each class were collected and 246 raw NIR spectrograms were obtained. 215 NIR spectrograms were utilised to establish the calibration model and rest 41 spectrograms were utilised to validate the performance of the model. On the original data, second derivative 9 point smoothing combined with the first range of calibration was chosen as the optimum pre-processing method at 3966~10414cm-1, and 30 solid excipients of all the 41 samples could be well identified by this model, namely “model 1”, and the other 5 sublibraries namely models(2-6) were developed for 11 excipients in order to classify every sample sufficiently. With this approach, no “error determinations” were assigned to a wrong class, which indicated the NIR method is appropriate for the identifacation of excipients in the cGMP manufacture of drug products within the pharmaceutical industry.