Mercury pollution is increasing both in the environment and in various organisms, especially top-predators. If variation in individual mercury concentrations is underpinned by genetic among-individual differences in traits related to mercury uptake, storage or excretion, and results in variation in fitness, populations may have the potential to evolve in response to this development. Few studies, however, have been able to collect sufficient information to investigate the genetic basis of pollution levels. We ran Bayesian quantitative genetic analyses, combining pedigree information obtained from a marine top-predator, the common tern (Sterna hirundo), with total mercury concentrations (THg) measured in 1364 blood and 1560 back feather samples obtained from >600 individual birds across seven years. Blood and back feather THg concentrations differed in repeatability (19 vs. 64%, respectively), but showed similar levels of heritability (c. 9%) and evolvability (c. 2%). Blood and back feather THg concentrations were positively correlated at the phenotypic, but not genetic, level. Although further work is needed to elucidate the uptake, storage or excretion pathways underlying these patterns, our study provides pioneering insights into the architecture of THg concentrations, and suggests that adaptation to environmental pollution may to some extent be possible.