In the design of offshore wind turbine generator (WTG) support structures, a large number of environmental parameters are used. The most important ones include the wind climate (wind speed and wind direction probability distributions), the wave climate (wave height and wave direction as well as peak frequencies of the waves, and water levels) and the soil conditions. These data in combination with the operational data of the wind turbine, such as the rotor speeds, pitch angle and yaw angle are used in dynamic simulations to determine the loads on the structure. Combining all these data correctly ensures that the support structure design, which is subject to large dynamic loads, is fitting for each specific location of the WTG. In particular, the environmental conditions should not coincide too much/often with the support structure natural frequencies, as that may result in resonance, thereby increasing fatigue life consumption. The dynamic behavior of these offshore support structures is of eminent importance. During the operation of WTGs, the actual dynamical properties of the WTG can be compared to the design. There are several reasons behind the goal of this comparison, for example verifying if the loading assumptions in the design were correct, assessing whether the life time can be extended or even whether the structure can be optimized (e.g. better foundations). Typically, this is done by comparing acceleration measurements of the WTG during idling (e.g. non-operational) conditions and applying Operational Modal Analysis algorithms. Given the high availability of modern wind turbines (e.g. >95% of time in operation) such measurements cannot be done as often as desired for this purpose. Therefore, the question arises to what extent the dynamical properties of WTGs can be estimated using measurement data obtained during operational (operational) conditions. This research explores the applicability of OMA algorithms in estimating the dynamical properties: the natural frequencies, the mode shapes and modal damping for operational conditions of offshore wind turbines on monopile foundations, using the Luchterduinen offshore wind park as a case study. The OMA technique explored in this thesis is Eigensystem Realization Algorithm. Test cases were used to confirm the performance of ERA for the estimation of the dynamic parameters. Stability diagrams were used to identify poles for stable modes (frequencies) in the frequency spectrum of data samples and to estimate a correct size of the Hankel matrices. It is concluded that this preprocessing and pre-analysis of the results is important to confirm the performance of ERA to correctly identify stable modes of support structures of operational offshore WTGs.Overall, it is concluded that ERA can yield useful results in estimating the natural frequencies for both ON and OFF, provided that an automated analysis is combined with a visual analysis. Even then, the accuracy and the consistency of the identified frequencies is moderate. ERA does not perform well for estimating the mode shapes, yielding inaccurate and inconsistent results. Based on the results of this work, ERA is found not to be capable of estimating modal damping ratios. The consistency for the mode shapes is not in all cases good and further study is recommended for both the mode shapes and the modal damping ratio. Another recommendation is to further study the size of Hankel matrices for this application. Furthermore, it is advised to also investigate the use of more advanced Operational Modal Analysis algorithms. The results of this study may be used to explore differences between wind turbines with and without scour protection and also to compare identified dynamic parameters to values used in the design.