This paper proposes a damage diagnosis strategy to detect and classify different type of damages in a laboratory offshore-fixed wind turbine model. The proposed method combines an accelerometer sensor network attached to the structure with a conceived algorithm based on principal component analysis (PCA) with quadratic discriminant analysis (QDA). The paradigm of structural health monitoring can be undertaken as a pattern recognition problem (comparison between the data collected from the healthy structure and the current structure to diagnose given a known excitation). However, in this work, as the strategy is designed for wind turbines, only the output data from the sensors is used but the excitation is assumed unknown (as in reality is provided by the wind). The proposed methodology is tested in an experimental laboratory tower modeling an offshore-fixed jacked-type wind turbine. The obtained results show the reliability of the proposed approach.