Abstract:Taken 83 varieties of wheats in Huang-huai area as material，screened target set by SPXY Algorithm，pretreated by Mahalanobis Distance Method，extracted sensitive wave points by SPA Method，and modeled by PLS Method，a near infrared spectral prediction model for wheat protein，wet gluten and sedimentation value of wheat was researched and built，and verified by using the prediction set.The results showed that the correlation coefficient（r）of the three results of protein content，wet gluten content and sedimentation value were 0.996 7，0.999 5 and 0.997 9，indicating that there was a good correlation and high credibility between the model and the conventional chemical values，which could provide reference for similar determination.