In this paper, a comprehensive survey for the state of art of a posteriori error estimates in finite element computation for magnetostatics since 1980s is presented.