To effectively obtain effective spectral response sub intervals of maize leaves under heavy metal lead pollution, and support heavy metal monitoring of crops. This article used hyperspectral remote sensing as the core technology and set up a maize pot experiment to collect a complete set of hyperspectral remote sensing data for maize leaves under heavy metal lead pollution using the SVC land cover spectrometer. A Lead Detection Index (LDI) was designed based on an improved Red Edge Normalization Index to obtain effective spectral response sub intervals of maize leaves under heavy metal lead pollution. Firstly, the original reflectance spectral data of maize in the training set was denoised by using the Db5 wavelet in the Daubechies wavelet series, resultingin the d5 component of the high-frequency component in the 5th layer of the wavelet decomposition. Then, we divided the entire spectral range of mazie leaves from 350 to 2 500 nm into 11 subband intervals and established LDI using the d5 wavelet coefficient values corresponding to the middle wavelength of each subband interval. Using the Pearson correlation coefficient r, LDI was compared with three conventional spectral indices (Photochemical Reflection Index, PRI; Meris Territorial Chlorophyll Index, MTCI; Modified Red Edge Simple Ratio Index, mSR). The effective spectral response sub-intervals of maize leaves under heavy metal lead pollution obtained from the training set data are purple valley, green peak, near-infrared platform, and near edge. The absolute values of Pearson correlation coefficients are all greater than 0.9, which are 0.911 0, 0.915 5, 0.905 1, and 0.907 6, respectively. In contrast, the absolute values of Pearson correlation coefficients between the three conventional spectral indices and the heavy metal lead content in the leaves are all less than 0.9, indicating high LDI effectiveness. Finally, we used the validation set data to obtain the effective spectral response subintervals of maize leaves under heavy metal lead pollution: purple valley, green peak, near-infrared platform, and near edge. The absolute Pearson correlation coefficients were all greater than 0.9. The Pears on correlation coefficients r for validation set one and validation set two were -0.999 9, -0.973 0, 0.914 2, 0.905 7, and -0.999 9, 0.911 7, -0.914 6, and 0.910 3, respectively. However, the absolute Pearson correlation coefficients between the three conventional spectral indices and the heavy metal lead content in the leaves were all less than 0.9. The results showed that under heavy metal lead pollution, the effective spectral response subbands of maize leaves were purple valley, green peak, near-infrared platform, and near-edge four subbands. The research results can provide technical support for monitoring heavy metal pollution in other crops.