Fusion proteins are an important class of proteins with diverse applications in biotechnology. They consist of 2 or more rigid domains joined by a flexible linker. Understanding the conformational space of fusion proteins conferred by the flexible linkers is important to predicting its behavior. In this paper, we introduce a modeling tool called FPMOD (Fusion Protein MODeller) which samples the conformational space of fusion proteins by treating all domains as rigid bodies and rotating each of them around their flexible linkers. As a demonstration, FPMOD was used to predict the fluorescence resonance energy transfer (FRET) efficiency of three different fusion protein biosensors. The simulation results of the FRET efficiency prediction were consistent with the in vitro experimental data, which verified that FPMOD is a valid tool to predicting the behavior of fusion proteins