Abstract: Using LOPEX’93 dataset, this paper simulated spectra reflectance, transmittance and absorptance of plant leaves under different biochemical constituents of leaf mesophyll structure parameter, water, chlorophyll, carotenoids and dry matter. The newest PROSPECT5 (leaf optical properties spectra) model was used for relational analysis of spectra and biochemical constituents as detected by vegetation remote sensing, on the basis of which red-edge characteristics were determined. The research task was to establish a model relating biochemical constituents to spectral characteristics via systematic relational analysis. Biochemical constituents of plant leaf should be inverted by the model. The results show that visible range of leaf spectra mainly depend on leaf mesophyll structure parameter and pigment (including chlorophyll and carotenoid). In the infrared range, leaf spectra largely depend on water, leaf mesophyll structure parameter and dry matter. Red-edge position moves towards long-wave with increasing chlorophyll, the so-called red-shift. Red-edge characteristics are not affected by water and carotenoid, and red-edge slope increases with increasing leaf mesophyll structure parameter. These relations are perfectly describable by di-order fitting equations. This lays the basis for solving mesophyll structure parameter issues.