Contaminated sediments pose a direct risk to sediment fauna and have the potential to affect other benthic assemblages. Disturbances that resuspend and remobilise contaminants may impact on filter-feeding, hard-substrate organisms that live immediately above sediments. This thesis uses laboratory and field manipulations to examine the impact of metal-contaminated sediments on sediment fauna and hard-substrate fauna simultaneously. It also compares the response of assemblages to metal contamination in a temperate and polar ecosystem. Simulated resuspension exposures in the laboratory indicated that contaminated sediments have the potential to affect hard-substrate organisms. Spirorbid polychaetes responded to both aqueous metals and to resuspended, particulate-bound metals. Impacts on hard-substrate fauna were however, not observed in manipulative field experiments using metal-spiked sediments. The recruitment and cover of hard-substrate organisms were either not affected or enhanced above contaminated sediments. In contrast, metal contamination had direct negative effects on sediment fauna, with a reduction in the abundance of most taxa. Results suggest that sediment fauna may interact with hard-substrate fauna through physical and/or biological mechanisms. In a reciprocal transplant experiment, established Antarctic hard-substrate assemblages were also unaffected by contaminant concentrations at an impacted site. Overall, metal-contaminated sediments are unlikely to pose as serious a threat to hard-substrate fauna as they do to sediment fauna. Contaminated sediments are not restricted to industrialised regions, and human activities in Antarctica have resulted in localised contamination near research stations. Although Antarctic assemblages are thought to be more sensitive than temperate assemblages to contaminants, few studies have explicitly examined this. Little evidence was found to support the theory that Antarctic assemblages are more susceptible to contaminated sediments. The response of Antarctic and temperate assemblages in the field to metal-contaminated sediments over a 10-11 month period was comparable. Responses were of a similar magnitude, despite differences in the composition of assemblages. In 10-d toxicity tests, the mortality of a common Antarctic hard-substrate organism was relatively insensitive to aqueous Cu, Zn and Pb. These results suggest that using current sediment quality guidelines from Australia may be a useful screening tool to assess the risk associated with contaminated sediments in Antarctica.