Abstarct: Abstract In this study, the seismic performance of a 5 MW offshore wind turbine founded on liquefiable soil with consideration of scour potential was investigated. Numerical modeling was carried out in OpenSees, incorporating different operational conditions of the turbine, namely parked, rated wind speed, and cut-out wind speed. All relevant loads, including dynamic aerodynamic wind loads using the BEM approach, static wind loads on the tower, and hydrodynamic loads baxsed on Morison’s equation, were modeled and applied to the structure. The soil profile consisted of a 50 m column with a 5 m loose sand laxyer overlying 45 m of dense sand, simulated for liquefaction using p-yLiq, t-zLiq, and Q-zSimple springs. Two scenarios, with and without 3 m scour depth, were considered to assess the impact of seabed condition on the seismic response of the turbine. Fourteen key geometric and mechanical parameters were selected for sensitivity analysis, with mean, upper, and lower bounds assigned for each. In total, 79 turbine models were developed and subjected to 80 real earthquake ground motions, resulting in 12,640 nonlinear time history analyses. Four engineering demand parameters (EDPs)—buckling, von Mises stress, RNA acceleration, and RNA rotation—were evaluated, and fragility curves were derived using the Box–Cox transformation and Bayesian inference for improved accuracy. The results indicated that excess pore water pressure accumulation prior to liquefaction follows the trend of Arias intensity, confirming its reliability as an indicator for liquefaction triggering. In terms of structural responses, buckling and rotation were most critical under the rated wind speed condition, whereas acceleration was most critical under the parked condition; this trend was consistent across both scour scenarios. Furthermore, scour increased all response parameters and consequently raised the probability of failure in all operational states.