The design of offshore wind farms is a complex process that requires a detailed study of the oceanographic, meteorological and geotechnical conditions at the site. The structure and all structural members shall be designed in a way that they can be resistant against different kinks of loads: permanent, variable, environmental, accidental and deformations. This paper is focused on those called environmental loads. The main environmental conditions that may contribute to structural damage, operational disturbances or other failures are wind, waves, currents and sea ice. Thus, the combination of the different parameters may produce many different critical situations for the integrity of the structure, requiring the calculation of long time series corresponding to long-term historical data situations. The most accurate techniques available at the moment to estimate loads acting upon a structure are numerical and physical models; however, they are very time consuming, and the calculation of long time series of data is unfeasible. Therefore, a new hybrid methodology to select waves–wind–current representative conditions that allow the interpolation of long time series of forces on a wind turbine is proposed. The methodology consists of a selection of a subset of representative cases of wave–wind–current climate at the structure's location by using a maximum dissimilarity algorithm, then estimating loads acting upon the structure for the sea–wind states selected and the reconstruction of loads corresponding to historical data using an interpolation technique based on radial basis function. To validate the proposed methodology and because of there is no availability of long time records of loads on wind turbines, the well-known IEC 61400–3 has been applied to estimate the loads for the complete reanalysis time series of waves, winds and currents. The validation of the results confirms the ability of the methodology developed to reconstruct time series of forces on the structure on the basis of the previously selected cases. This methodology permits application of numerical and physical models to offshore wind farm design, considerably reducing the number of tests or simulations. Copyright © 2012 John Wiley & Sons, Ltd.