Trophic models based on nitrogen stable isotope ratios (d15N) have been shown to predict changes in mercury (Hg) concentrations in fish; however, they are usually applied at the ecosystem scale and are rarely directed at known trophic pathways. We discuss a novel approach in which we combined gut contents analysis and stable isotope analyses (d15N and d13C) into a Bayesian isotopic mixing model to provide a quantitative estimate of Hg and selenium (Se) biomagnification in an estuarine food web. Estimates of the relationship between total mercury (THg) and methylmercury (MeHg) were significantly improved in mixing model-adjusted food webs over models that included all known prey sources. Spatial differences in dietary composition and MeHg bioavailability offer strong evidence that local food webs can have a significant effect on the biomagnification of Hg within benthic fish species. While no evidence of Se biomagnification was found, lower Se :Hg ratios at higher trophic levels could be attributed to increasing trophic Hg concentration. Furthermore, stable isotope analysis suggested Hg and Se biotransfer from benthic sources to fish. Overall, the findings highlight that isotope mixing models can be a significant aid in assessments of contaminant biomagnification, particularly when it is important to define food pathways to top predators. © 2014, by the Association for the Sciences of Limnology and Oceanography, Inc.