In resource assessment, an important consideration is the interval over which to average the measured wind speed before storing the sample in the data logger. One choice is to use the sample interval that was used to measure the powercurve: 10’ for large wind turbines, 1’ for small ones. This, however, is not always possible, e.g. when data from a meteorological station are used. These are typically averaged and stored for every hour.In this paper we study the error in average energy production estimates when an inconsistent sample interval is used to store wind speed measurements. Based on wind speed measurements of one onshore and one offshore site and an independently-tested power curve, we estimate the mean power and show how it varies when the sample interval changes. We also illustrate how the power curve can be compensated for the inconsistent sample interval (of the wind speed measurement), by exploiting the correspondence between this effect and the effect of turbulence.We conclude that the error on the average energy production can be reduced and suggest how this can be realised in a practical wind resource assessment campaign.