Abstract: At present，the non-destructive detection technology of wheat biomass has problems of low precision and low efficiency.Mechanical sensing technology and multiple regression analysis were used to construct prediction model of wheat aboveground fresh biomass，and prediction effect of wheat biomass in different growth stages was explored.Results showed that prediction effect of the model was the best when measurement height was 40 cm；prediction effect of the model was the best when measurement height was 40 cm.R2 value of the test set was 0.623，and RMSE value was 0.216 kg.Combination of measurement height and plant height characteristic parameters could not significantly improve prediction effect of wheat biomass.This study has a certain reference value in the rapid and non-destructive detection of wheat breeding.