A comprehensive compilation of the published data for trace element concentrations (metals and metalloids) in Australian marine mammals, birds and turtles is presented. The majority of studies have relied on the utilisation of opportunistically collected samples, animal strandings and bycatch. This has resulted in large gaps in geographical, temporal and species coverage data. For instance, little or no data are available for cetaceans in New South Wales or the Northern Territory, and out of 14 endemic species of dolphins, data only exist for seven species. The aforementioned data gaps make it hard to identify statistically significant trends, a problem compounded by data being reported in the form of ranges without raw data. Trace element concentrations measured in various marine species and their tissue types are extremely variable, with ranges typically spanning several orders of magnitude, but are generally comparable with international data. Trends in contaminant concentrations with tissue type follow generally accepted patterns of behaviour for higher organisms, with the highest mercury concentrations in liver and cadmium in kidney tissues. Herbivores have lower contaminant loadings than carnivores, reflecting the importance of diet, and there are identifiable age-related trends for elements such as mercury. The lack of supporting pathology on dead and stranded animals and data on specimens from uncontaminated locations restrict conclusions on organism health impacts. There have been some attempts to use non-invasive sampling of indicator tissues such as fur, bristle and feathers. However, it is currently difficult to extrapolate these data to estimate contaminant concentrations in major organs. Recommendations for future investigations are made.