The effect of heavy metal copper stress on maize leaves reflectance is weak, hidden and difficult to be detected. To analyze the spectral response of maize leaves to copper stress and discriminate the stress degree on maize plant, the pot experiment with 11 copper stress gradients was set at ground laboratory, the maize leaves reflectance and chlorophyll content were obtained at seedling stage, jointing stage and heading stage. The copper content in leaves was also gotten at heading stage. At 480~670nm and 670~750nm, the spectra simulated under the conditions of normal growth, low concentration stress and high concentration stress were constructed respectively. The mechanism of using the first three harmonic amplitudes C1, C2 and C3 to detect the weak spectral distortion in 〖JP2〗maize leaves reflectance under copper stress was analyzed. The harmonic analysis was carried out in two wavelength ranges of 480~〖JP〗670nm and 670~750nm at seedling stage, jointing stage and heading stage, and the rule between the amplitudes C1, C2 and C3 and the copper stress gradient was also analyzed. The results showed that at the seedling stage, in the gradient range of Cu(100) to Cu(1200), with the increase of copper stress gradient, the harmonic amplitude C1 of old leaves reflectance of 480~670nm and 670~750nm was increased, and in the gradient range of Cu(100) to Cu(1200), the harmonic amplitude C1 could be used to distinguish the degree of copper stress on maize plant. At heading stage, from Cu(50) to Cu(1200), except for Cu(1000) gradient, the harmonic amplitude C1 of 480~670nm was increased with copper stress gradient. At jointing stage, for the bands of 480~670nm and 670~750nm, the variation of the harmonic amplitudes C1, C2 and C3 was not obvious in all stress gradients. The seedling stage and heading stage were the best growth stages for monitoring copper stress on maize plants by harmonic amplitude characteristics. The method proposed was designed and tested at leaf scale, and further study were also needed to evaluate its potential at canopy scale. The method proposed provided a potential basis for other applications by using airborne and satellite remote sensing data.