Monopile-supported offshore wind turbines (MSOWT) are constantly subjected to various cycles of forces, including wind, waves, currents, and seismic activity. To avoid resonance problems, it is crucial to ensure that the structure’s natural frequencies are distinct from those of these external forces. However, calculating these natural frequencies can be computationally expensive, especially considering the dynamic interactions between fluids, soil, and structure. To address this, an artificial neural networks (ANN) based surrogate model is proposed to estimate the eigen frequencies of the MSOWT. The comprehensive procedure for generating the training dataset using the Latin hypercube sampling (LHS) method is detailed. The output and input parameters are related using a rectified linear unit (ReLU) activation function based on Pearson correlations. The results show that this model can predict natural frequencies with over 90% accuracy. The surrogate model is applicable for multiple evaluations having high level of precision, combined with a significant reduction in computational costs.