The objective of this paper is to present a system identification method suitable for miniature rotorcrafts under hovering. The proposed model to be identified is a Takagi-Sugeno fuzzy system, representing translational and rotational velocity dynamics. For parameter estimation of the Takagi-Sugeno system a classical Recursive Least Squares (RLS) algorithm is used, which allows identification to take place on-line since parameter updates are produced whenever a new measurement becomes available. The validity of this approach is tested using the X-Plane ^© flight simulator. Data obtained offer justification for the applicability of the approach in real-time applications.