Helicopter flight simulators require high-fidelity dynamic models based on a set of stability and control derivatives (SCDs). For different flight conditions, specific sets of SCDs must be identified and utilized in the flight simulator software. In this study, nine sets of SCDs of the AS 355 F2 helicopter have been estimated from experimental flight-testing data, considering different combinations of velocity, mass, and altitude. Linear dynamic models with eight states and four controls were determined for the different flight conditions, using the output-error method. The associated optimization problem was solved with the Gauss–Newton algorithm, and identification reliability was evaluated by the Cramer–Rao (CR) and relative correlation coefficient (RCC) indices. An instrumented AS 355 F2 Helicopter belonging to the Instituto de Pesquisas e Ensaios em Voo, Departamento de Ciencia e Tecnologia Aeroespacial, Sao Jose dos Campos, Brazil, was utilized for data collection in different flight conditions, totaling 16 test hours. The following natural dynamic modes have been excited: phugoid, short period, spiral and Dutch roll, by command inputs of the type: frequency sweep (sinusoidal), doublet pulse and 3-2-1-1. Considering CR and RCC criteria, the most appropriate maneuver for exciting each dynamic mode is suggested, and the sets of the nine flight conditions are SCDs reported.