A field experiment investigating the effect of oil contamination on benthic microbial communities was conducted near Casey Station, East Antarctica. Defaunated sediment was treated with a mixture of Special Antarctic Blend diesel and lubricating oil and deployed in three different bays for eleven weeks. A molecular fingerprinting technique, denaturing gradient gel electrophoresis (DGGE), was used to investigate the microbial community structure. The variation between replicate samples within treatment groups indicates that the benthic microbial populations are very diverse and evenly distributed. Comparisons to determine the significance of both deployment location and hydrocarbon treatment showed that the greatest effect was from a combination of location and treatment. Detailed analysis suggests that subtle differences may be obscured by variability introduced by PCR and gel stages in DGGE, undermining this experimental approach. It is concluded that both location and hydrocarbon contamination influenced the development of the microbial communities but that the effect of hydrocarbon treatment varied with location. This has important implications for the design of future experiments on the effect of hydrocarbons on benthic communities, especially if it is intended to generalize the conclusions drawn from site specific studies.