The complex geometry of the human lung airway tree makes the airflow simulation of the whole tree unreachable. As a first difficulty, the resolution of CT imaging is currently restricted, and smaller tubes are not visible during segmentation of the 3D models. Secondly, it is not computationally feasible to simulate all generations in any case. Thus most of modeling is restricted to the simple geometries or only a set of generations. The aim of this study is to obtain a model that could accurately describe the airflow in the whole respiratory track. In this paper a multiscale modeling of the respiratory track is proposed by using Monte-Carlo method to construct a stochastic structure of the lung. In this way the respiratory tree is decomposed into three regions and a different model will be exploited in each of these regions: the upper part (before lobes entrances) where the unsteady incompressible Navier-Stockes equations hold to describe the fluid flow in 3D coordinates, the distal part (inside lobes before acinar region), where one can assume one dimensional flow and the acinar airways assumed to be embedded in a box and a lumped parameter model based on Womersley analysis can be used. We have applied the lumped parameter model of acinar region as the outlet boundary condition of the distal part. According to this model the impedances are computed at the end of the distal part. Then for each tube based on the ratio of the pressure drop to the flow rate in the frequency domain, an impedance is introduced. Starting from the acinus and moving upward, the impedance at each tube inlet will be approximated. By calculating these impedances we can find the flow rate and outlet pressure of each tube. After solving the 1D Navier-Stockes equations, the tubes impedances are corrected until the convergence is achieved. One important issue is coupling between 3D and 1D model. This goal can be achieved by using the following algorithm. Starting from a given inlet velocity and an estimation for the pressure at the outlet, the 3D model is solved. Using the computed flow rate at the outlet of this model, the 1D model is solved in the next step. Then the computed inlet pressure of 1D model is compared with the initially assumed outlet pressure of 3D model and in case of any difference the pressure is corrected, accordingly. The procedure is repeated until coupling is established within a tolerance. This algorithm is used in each time step to achieve complete coupling. Finally regional and total deposition fractions of different particles in the human respiratory system during inhalation are computed using available semi experimental equations and stochastic algorithm. Keywords: Monte Carlo modeling, Whole lung, Five lobes, Impedance, 3D-1D coupling, lumped parameter model, particle deposition