The aim of this doctoral thesis was to investigate and improve the risk assessment of anti-fouling paints. A new method, the Petri-dish method, was developed to determine release rates of copper and zinc from anti-fouling paints (Paper I). The release rates of zinc were substantially higher from the biocide-free leisure boat paints than from the biocide-leaching paints. In Paper II, the potential toxicity of paint leachates was assessed and the biocide-free paint proved to be the most toxic paint investigated. Zinc, a supposedly non-biocidal ingredient in copper-based and other antifouling paints, was found to contribute significantly to the observed toxicity of all leisure boat paints investigated. This means that risk assessment of an anti-fouling paint based on only the active biocidal ingredient in the formula is insufficient. A more holistic approach, based on hazard identification and dose-response assessment of anti-fouling paint leachate, with all ingredients taken into consideration, is recommended. This can be achieved by the Petri-dish method, which combines chemical analysis with ecotoxicological tests of paint leachates. In Paper III and IV, the effects of salinity and organic matter on copper bioaccumulation and toxicity to the red macroalga Ceramium tenuicorne were studied. Salinity had only a minor effect in ameliorating copper toxicity, whereas the organic matter concentration had a significant effect in reducing the bioavailability and hence copper toxicity at all salinities tested. Copper uptake and bioaccumulation by C. tenuicorne showed that the macroalga could access a sizeable fraction of organically-complexed copper in addition to Cu²⁺, when Cu²⁺ concentration to the cell membrane is diffusion limited. This observation implies that the setting of environmental quality standards (EQSs) for copper and other metals through the Biotic Ligand Model (BLM) is inappropriate in predicting copper uptake and hence toxicity to C. tenuicorne.