Abstract: For greenhouse crop system, significant differences in response times exist. Crop dry matter production responds rather slowly to the changes in the environmental conditions while the greenhouse temperature responds relatively fast. For this reason, a greenhouse temperature hierarchical control system was proposed. The hierarchical system was designed with two layers, including the crop layer and the climate layer. First, in the crop layer, a decoupled lower order model was established based on the theory of singularly perturbed systems. The controller was handled by means of an optimal control approach to achieve temperature set points, minimize the net cost according to objective function. Then, a fuzzy controller in the climate layer was designed to cope with the problem of tracking set points obtained from the crop layer. At last, the designed hierarchical control system was tested in Matlab. The simulation results show that both the crop and climate layer controllers can work well with better control performance.