Fidelity is important because it is at the heart of what distinguishes computer simulations from any other computer program. The development of distributed simulations causes the desire for new and more comprehensive fidelity methods to describe and quantify fidelity. First, the current status of study on fidelity is analyzed. Second, fidelity theories are discussed, the theorem, definition, description, and metrics of fidelity are presented, and these theories are the base for the study on fidelity. Finally fidelity management process (FiMP) is presented, and this process can validly guide the development and evaluation of fidelity.