This paper presents a two-stage method that identifies scale-model helicopter dynamics without requiring to have good initial values of the parameters to be estimated. Stage one is a pre-estimation process that determines good initial values to start stage two. The novelty demonstrated in this paper is that this new initial value determination method is efficient and methodical compared to the cumbersome and tedious experiments and/or mathemati- cal computations required to obtain approximate initial values. The pre-estimation process then delivers high quality initial values that can be used in the second stage of the esti- mation process. As the second stage the frequency-domain output-error based parameter estimation process is used. The pre-estimation stage uses the partitioned subsystems of the helicopter model and the final estimation process uses the entire helicopter model as one system that describes the hover and near hover linearized dynamics. All flight data used through out this paper are real flight data. The model validation and prediction results are presented to show the validity of the method. © 2012 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.