Various loads induced by marine environmental conditions, such as waves, currents, and wind, are crucial for the operation and viability of offshore wind power (OWP) systems. In particular, waves have a significant impact on the stress and fatigue load of offshore structures, and highly reliable design parameters should be derived through extreme value analysis (EVA) techniques. In this study, extreme wave analyses were conducted with various Weibull distribution models to determine the reliable design parameters of an OWP system suitable for the Ulsan area. Forty-three years of long-term hindcast data generated by a numerical wave model were adopted as the analyses data, and the least-squares method was used to estimate the parameters of the distribution function for EVA. The inverse first-order reliability method was employed as the EVA technique. The obtained results were compared among themselves under the assumption that the marginal probability distributions were 2p, 3p, and exponentiated Weibull distributions.