 This study compared the accuracy of wind speed predictions from three international re-analysis datasets, BARRA, ERA-5, and MARRA-2, for offshore wind farms in Australia. The results showed that the BARRA dataset had the highest correlation and lowest average error when compared to weather station data. Additionally, significant variation was observed between the datasets, indicating the need for additional site-based measurements. This article was authored by Emily Cowan, Chonglong Wang, and Stuart D. C. Walsh.