In this paper, the authors propose the novel comprehensive probabilistic analytical model of the sub-threshold MOSFET's (Metal-Oxide Semi-conductor Field-Effect Transistor) performance affected by both random dopant fluctuation and process variation effects. The up to dated Takeuchi's physical level random variation model has been adopted. The proposed model has been found to be analytic, powerful and comprehensive as it has been derived by using the sub-threshold MOSFET's physical equation without any approximation. This model has been verified at the nanometer level i.e. 65nm CMOS process, by using the BSIM4 based (MC) Monte-Carlo simulations.