The invention discloses a canopy hyper-spectral triticum aestivum leaf dry weight monitoring method based on continuous wavelet analysis. The method comprises the following steps: selecting sampling plots, acquiring the triticum aestivum canopy hyper-spectral reflectance, and measuring the triticum aestivum leaf dry weight, wherein the sampling plots are selected from different test points, are of different varieties, different nitrogen levels and different planting density, and are from different years; carrying out continuous wavelet transform on the acquired triticum aestivum canopy hyper-spectral reflectance data to get a wavelet coefficient under a specific wavelength and a specific scale; using the wavelet coefficient to analyze the quantitative relation between triticum aestivum leaf dry weight and wavelet coefficient, screening out an optimal wavelet function sensitive to triticum aestivum leaf dry weight and a characteristic value corresponding to the optimal wavelet function, and building a triticum aestivum leaf dry weight quantitative model based on continuous wavelet analysis; and using independent triticum aestivum test data to assess the reliability and applicability of the quantitative model, and using a determination coefficient R<2> and relative root mean square error RRMSE between predicted and observed values to evaluate the quantitative model.