The recent discovery of mass-independent fractionation of mercury isotopes allows new constraints to be placed on the mercury cycle. Here we report new Hg isotopic analyses of zooplankton and fish from different trophic levels of a freshwater lake (Lake Jackson, Florida) bearing systematic mass-independent fractionation of mercury isotopes. Fish muscle tissues show a progressive enrichment in the odd-mass mercury isotopes having odd atomic mass numbers (199 and 201) with increasing trophic level. Trophic level was determined based on nitrogen isotopic composition as well as fish stomach content. Zooplankton in the lake contain mercury with Δ¹⁹⁹Hg and Δ²⁰¹Hg values of +0.43 (±0.07)‰ and +0.44 (±0.07)‰, respectively. The Δ¹⁹⁹Hg values increase by ∼1‰ from ∼+0.4‰ in zooplankton, juvenile bluegill, and several other small fishes to Δ¹⁹⁹Hg = +1.36‰ for the Florida gar, which is the top predator fish in the lake. Previous observations of odd-mass-number isotope enrichment of mercury have been explained by photoreduction and demethylation of methyl mercury in the water column or as isotope effects related to microbial methylation. However, our data and the data of Jackson et al. (2008) are also consistent with in vivo production of mass-independent fractionation. Considering the alternatives, mass-independent fractionation by metabolic processes is the most straightforward explanation for the mercury isotope data. There are two known mechanisms for mass-independent fractionation of mercury, i.e., the nuclear volume effect and the magnetic isotope effect. While the data are insufficient to serve as proof, the magnitude of the mass-independent effect and the nearly equal enrichment of ¹⁹⁹Hg and ²⁰¹Hg seem most suggestive of a magnetic isotope effect.