The UK, like many other industrialised countries, is committed to reducing greenhouse gas emissions under the Kyoto Protocol. To achieve this goal the UK is increasingly turning towards wind power as a source of emissions free energy. However, the variable nature of wind power generation makes it an unreliable energy source, especially at higher rates of penetration. Likewise the aim of this paper is to measure the potential reduction in wind power variability that could be realised as a result of geographically dispersing the location of wind farm sites. To achieve this aim wind speed data will be used to simulate two scenarios. The first scenario involves locating a total of 2.7 gigawatts (GW) of wind power capacity in a single location within the UK while the second scenario consists of sharing the same amount of capacity amongst four different locations. A risk portfolio approach as used in financial appraisals is then applied in the second scenario to decide upon the allocation of wind power capacity, amongst the four wind farm sites, that succeeds in minimising overall variability for a given level of wind power generation. The findings of this paper indicate that reductions in the order of 36% in wind power variability are possible as a result of distributing wind power capacity. (c) 2007 Elsevier Ltd. All rights reserved.