Field data of benthic communities and contaminant loadings in marine sediments measured in parallel can be used to derive sediment quality guidelines (SQGs) using a field-based species sensitivity distribution (f-SSD) approach. Recently, SQGs have been successfully derived from f-SSDs for the Norwegian continental shelf with an extensive survey (>1 million km2) and a large data set (1,902 sampling stations with 1,944 species). The present study examined the practicality of this approach in deriving SQGs for a much smaller geographical area, namely, the marine environment of Hong Kong (sea area: 1,651 km2), making use of databases of the government of Hong Kong special administrative region. As the construction of f-SSDs requires the use of a collection of responses from individual species to a chemical gradient in sediment, data screening criteria on the minimum abundance of the species were evaluated and optimized to ensure sufficient statistical power for estimating these responses. Sediment quality guidelines were derived for nine trace metals, total polycyclic aromatic hydrocarbons, and total polychlorinated biphenyls and compared with current SQGs in developed countries. The community-adjusted hazardous concentrations of 5% and 10% of the f-SSDs were adopted to represent the threshold effects level (TEL) and predicted effects level (PEL), respectively. The TELs derived from this f-SSD approach compares favorably with current SQGs, while the derived PELs were generally lower than the current SQGs, indicating that they are more protective. The f-SSDs can be directly utilized for probabilistic risk assessment, while the field-based SQGs can be used as site-specific guidelines or integrated into current SQGs. Our results suggest that the f-SSD approach can also be applicable to small areas such as Hong Kong.