Potential wind power for a given period (e.g. a day) can be determined from wind speed data measured in certain hours of a period. Obviously, the sum of the cubes of wind speeds measured depends on the number of measurements. This dependence can be reduced in two ways: determining the average and the relative wind energy for a given time within a given period. The method of sliding averages uses both. Applying this method a given hourly average wind speed cube of a day is estimated on the basis of wind speeds measured in that hour of the day. Cubes of the wind speeds are in proportion with the total daily potential and produced wind energy. This model requires long-time series of wind speed data that are available only for weather stations in Hungary, where hourly average winds speeds are registered. For this reason, statistics required for the model were calculated from different subsets of ten-yearlong hourly average wind speed time series of three Hungarian weather stations (Szombathely, Budapest-Lorinc and Debrecen). Using the statistics and hourly wind speed data measured in the vicinity of the wind turbines/on the wind turbines themselves, the model is suitable for giving estimations hourly of the potential wind energy for the whole day in a particular season or circulation type group. A software for the model is also presented here. Considering the results the sliding average model (SLIDAV) makes it possible to forecast average daily wind power 6-9 h before the end of the day with an error of 20%. The magnitude of the error of estimation depends on the given season and/or synoptic type group. These results may provide important information for wind turbine owners: daily amount of wind energy can be determined in this way. Thus the owner can decide whether to operate the turbine whole day, or to stop it periodically for maintenance for example. (C) 2010 Elsevier Ltd. All rights reserved.