This article describes a new parametric method for the development of nonlinear models with parameters identified from an experimental setting. The approach is based on applying a strong non-resonant high-frequency harmonic excitation to the unknown nonlinear system and monitoring its influence on the slow modulation of the system’s response. In particular, it is observed that the high-frequency excitation induces a shift in the slow-modulation frequency and a static bias in the mean of the dynamic response. Such changes can be directly related to the amplitude and frequency of the strong excitation offering a unique methodology to identify the unknown nonlinear parameters. The proposed technique is implemented to identify the nonlinear restoring-force coefficients of three experimental systems. Results demonstrate that this technique is capable of identifying the nonlinear parameters with relatively good accuracy.