A two-dimensional dispersion and consequence model called RADIS has been developed and used to perform a site-specific investigation of the consequences and risks for Hong Kong as a result of a potential accident at the Daya Bay Nuclear Power Station. The study uses the 1986 Hong Kong weather and demographic data to determine the accident consequences for Hong Kong. The overall cumulative occurrence frequency distributions of early fatality, early injury, latent cancer, and thyroid nodule are compared with those of WASH-1400 and a United Kingdom Atomic Energy Authority (UKAEA) study. This study is generally consistent with the WASH-1400 results and exhibits less conservatism than the UKAEA study.