Assessment of the structural integrity of an offshore wind turbine requires simulation of its structural response for many combinations of the wave and wind conditions that represent its lifetime loads. Modeling the dependence between the wind speed, wave height and wave period is very expensive because it requires a large amount of data. Assuming independence can lead to inaccurate estimation of the probability of failure. Some researchers assume that the wave height follows a standard distribution conditioned upon wind speed. We propose an alternative method that uses a copula to approximate the joint probability distribution of the wind speed, and the significant wave height. We believe that this approach is more complete because we obtain the joint distribution of the above quantities without making any assumption on their conditional distributions. To test the quality of the proposed approach, we use Monte Carlo simulation to generate sample values of the wind speed and wave heights, and validate the model with the available data.