Evaluating the structural robustness of an onshore wind turbine tower necessitates the quantification of its stochastic dynamic behavior subjected to variable wind loads. Nevertheless, the intricate nonlinearity inherent in the turbine’s governing dynamics poses a substantial computational challenge for time-domain analysis of stochastic responses. To mitigate this, we introduce a frequency-domain linearization approach for the tower’s dynamic modeling, thereby streamlining the computation process. This research presents a frequency-based framework for modeling wind turbines, leveraging an analytical technique to derive the power spectral density (PSD) of wind velocity and the corresponding loads on the rotating blades. A flexible multi-body dynamics (MBD) framework is adopted to construct the dynamic model of the onshore turbine, where the inherently nonlinear dynamic equations are linearized into a time-invariant form through a multi-blade coordinate (MBC) transformation. Following this, the PSD of tower displacements and internal forces are computed within the frequency domain. The analytical findings align closely with Monte Carlo simulation (MCS) outcomes, with a maximum relative discrepancy of 8.4% in statistical metrics (mean and variance), underscoring the efficacy and accuracy of the proposed frequency-domain methodology.