In the classical modal parameter estimation approach, the baseline data which are processed are Frequency Response Functions measured in laboratory conditions. However, in many applications, the real operating conditions may differ significantly from those applied during the modal test (suspension pre-strains of a car on the road, large motion of a running engine, aeroelastic interaction and wing/rotor pre-strain in air- and rotorcraft, strong environmental excitation in civil structures as traffic load, wind/wave excitation,.... ). Hence, the need arises to identify a modal model in these real operational conditions. In most cases, only response data are measurable while the actual loading conditions are unknown. A number of different estimation techniques have been evaluated for their applicability in such conditions. They include the Polyreference Least Squares Complex Exponential method and an Auto-Regressive Vector model method. The basic principles of these methods are reviewed and applied to an experimental case.