The assessment of the power performance of a wave energy converter (WEC) at a given site involves two tasks: (i) the characterisation of the wave resource at the site in question, and (ii) the computation of its power performance. These tasks are generally seen as disconnected, and tackled as such; they are, however, deeply interrelated - so much so that they should be treated as two phases of the same procedure. Indeed, beyond the characterisation of the wave resource of a certain area lies a crucial question: how much power would a WEC installed in that area output to the network? This work has two main objectives. First, to develop a methodology that integrates both tasks seamlessly and guarantees the accurate computation of the power performance of a WEC installed at a site of interest; it involves a large dataset of deepwater records and the implementation of a high-resolution, nested spectral model, which is used to propagate 95% of the total offshore wave energy to the WEC site. The second objective is to illustrate this methodology with a case study: an Oscillating Water Column (OWC) projected to be constructed at the breakwater of A Guarda (NW Spain). It is found that the approach presented allows to accurately determine the power that the WEC will output to the network, and that this power exhibits a significant monthly variability, so an estimate of the energy production based on mean annual values may be misleading. (C) 2012 Elsevier Ltd. All rights reserved.