The paper uses Near Infrared Spectroscopy to detect the quality of wet gluten in wheat flour. Three pretreatments including eliminating the constant offset、multiplicative scatter correction and less a straight line are used to deal with the near infrared spectra，then create the related model between the near infrared spectrum of wheat flour and wet gluten content using the partial least squares method, then predict the wet gluten content in wheat flour. Experimental results show that the evaluation of model is optimal when using less a straight line to pretreat the near infrared spectrum，and the correlation coefficient R2 is 934 and the RMSECV is 051 after cross validation, the root mean square error of predicted results is 0456. Experimental results show that near infrared spectroscopy can be used to determine the quality of wheat flour gluten content.
