Herein, we propose an imaging method based on frequency modulation and spatial coding to overcome the problem associated with fluorescence molecular tomography data acquisition methods, improve the data acquisition scheme, and reduce data acquisition time. In this method, the excitation beam is split into several sub-beams and used as multipoint excitation source. These sub-beams are modulated to different frequencies and then simultaneously incident at different points on the target surface. At the detection side, the emergent light of the target is first directed to a single photomultiplier through a spatially coded mask. According to the compressed sensing theory, the distribution of fluorescence signals on the target surface can be obtained by changing the mask mode and conducting sparse reconstruction and recovery. We design the corresponding simulation experiment to evidence the proposed methodology. Result shows that the method can better restore the original image, which proves the feasibility of the proposed method.