Particle transport in the human upper respiratory tract using large eddy simulation (LES) for flow solution involves a time consuming dynamic procedure that continuously couples the flow and the particle calculations. This paper focuses on an alternative efficient multiple LES frozen field approach for particle transport solution in the human upper respiratory tract. The methodology allows decoupling the particle simulation from the LES simulation, resulting in important saving on computational time, if different particle sizes are to be considered. A systematic procedure is presented for determining an optimal set with several instantaneous LES frozen fields. The sampling period and the time interval between frozen fields are the two base parameters for the optimal set determination. These parameters are evaluated based on a systematic analysis of eigenvalues and eigenmodes obtained for various sample sets of LES flow field data bank by using the discrete Proper Orthogonal Decomposition method (POD). In our upper airway geometry, the present procedure led to a good agreement with the corresponding experimental particle deposition data, and significantly reduced the resources required for a full LES approach for flow and particles. (C) 2013 Elsevier Ltd. All rights reserved.