Adaptive time-stepping is a method of choosing a variable time-step size for time-marching applications given prescribed tolerance for local error. When applied to steady-state problems this method acts as a convergence accelerator while for transient problems it maintains the time accuracy using a user-prescribed error criterion. In this paper, we demonstrate the usefulness of this method over the commonly-used CFL based time-stepping for different problems with steady/unsteady laminar viscous flow. © ICCFD 2018.