Observation and assessment of wind resources is a prerequisite for wind farm construction. Due to the investment cost of offshore wind farm is very expensive, more accurate assessment of wind resources is needed to reduce their investment risks. From traditional field observation to multi-platform remote sensing and from ordinary mathematical statistics to coupled numerical model simulation, abundant offshore wind data and evolving assessment methods make the results of offshore wind resource assessment more and more reliable. Poor station observations and rich remote sensing data are distinct characteristics of offshore wind data. Technology integration of applying multi-scale coupled models to assimilate multi-source remote sensing and station data is a mainstream development direction of offshore wind resource assessment methods. The wind resource assessment for offshore wind farm should focus on data quality and method selections of data interpolation, wind speed calculation of return period and wind energy parameters adjusted for a long term condition because these factors can significantly affect the operating efficiency of future wind farm.