||Hong Kong has mainly relied on imported fossil fuels to support its energy system, and such burning of fossil fuels to generate electricity has recently raised public awareness of energy efficiency and environment issues. There is, however, the potential to develop wind power in Hong Kong, since most of the land in Hong Kong is hilly, with a complex coastline and numerous islands, offering many likely sites. The main objectives of the study are to advance understanding of the characteristics of the wind flow field over hilly complex terrain near the costal region of Hong Kong, in order to explore the potential for use in wind energy applications. This thesis includes analysis of wind speed data for six typical sites in Hong Kong, in terms of site terrain, wind conditions and wind power potential. Selected locations, including mountain tops and outlying islands wind power resources, are assessed for annual and monthly wind speed and power densities. In order to increase understanding of the characteristics of the wind flow field over hilly complex terrain for use in wind energy applications, the boundary layer wind tunnel at the CLP Power Wind/Wave Tunnel Facility (WWTF) is used to study the physical wind flow over the topography of Hong Kong by conducting topographical wind studies, using 1:2000 scale models of Hong Kong. Measurements of mean flow and turbulence parameters for six different directions and various heights near the crest of a hill top at Yi Tung Shan (YTS), 750 m high, are presented. The most striking features of the longitudinal mean velocity are the increases in velocity near the surface on top of the hill, the slight deceleration at the upwind hill foot, and the flow separation behind the hill. The speed-up effect can be determined through a wind tunnel topographical study used to calculate the topographical multiplier (T.M) at a particular point. For YTS, heights of 40 to 60 m for wind turbines appear to be adequate based on the results from the topography study. The best prevailing wind direction can be also obtained. The final goal of this study was an investigation of the capability of physical modeling of Hong Kong topography to predict wind speed distributions by comparing predictions with full-scale data. The ratios of the mean wind speeds in the wind tunnel at locations corresponding to two full-scale anemometer positions were obtained for all wind directions and were used to transfer data from one site to the other. However, agreement between the predicted and actual wind speed distributions for the hilly sites were not obtained initially, when only the wind speed ratios obtained from the Chek Lap Kok (CLK) reference site, with a 1:2000 scale model, were used. It was found necessary to measure the velocity profiles through one of the anemometer positions close to the terrain using a detailed 1:400 scale Waglan Island (WI) model, and use these profiles to correct the 1:2000 velocity ratios for local terrain effects in order for accurate predictions to be made. This investigation enables a classification of the classes of wind power density that can be predicted at a site without the general wind characteristics information, to be made. The wind tunnel studies provided very useful information on the characteristics of the wind flow field over hilly complex terrain, and the results were integrated with the wind speed statistics collected at nearby meteorological stations to estimate the wind energy budget, with which the potential of wind power generation can be assessed.