Detecting anthropogenic metal contamination in regional surveys can be particularly difficult when there is a lack of pre-disturbance data, especially when trying to differentiate low to moderate levels of contamination from background values. Furthermore, comparisons with other regional studies are confounded by differing analytical methods used and variations in sediment properties such as grainsize. Several types of geochemical technique, including weak acid partial extraction, strong acid extractions and total digestion have been used. Attempts have been made to overcome the influence that grainsize has on chemical concentrations in heterogeneous environments by analysing the fines, typically the mud fraction (<63 μm), in an attempt to improve the detection of anthropogenic contamination. Here we compare a weak acid partial extraction using 1 M HCl and total digestion methods for a regional survey of reference and impacted sites in Antarctica using both whole sediment (<2 mm) and mud (<63 μm) fractions. The 1 M partial extraction on whole sediment (<2 mm) most closely distinguished weakly, or moderately, impacted sites from reference locations. It also identified small scale within-location spatial variation in metal contamination that the total digest did not detect. Compared with total digests or analysis of the <63 μm fraction alone, this method minimised the possibility of a Type II statistical error in the regional survey – that is, failing to identify a site as being contaminated when it has elevated metal concentrations. To allow inter-regional comparison of sediment chemistry data from elsewhere in Antarctica, and also more generally, we recommend a 1 M HCl partial extraction on whole sediment (<2 mm).