In December 2007, a mass mortality of isaza (Gymnogobius isaza), a goby fish in Lake Biwa, Japan, was observed under severe hypoxia. Considering the level of manganese and arsenic in the dead isaza during the event was much higher than that in live isaza, hypoxia-induced mobilization of manganese and arsenic and subsequent exposure could be the reason for this adverse effect. However, secondary accumulation of manganese and arsenic after the mortality event could not be ruled out. To test this hypothesis, we conducted tissue distribution/speciation analysis and absorption tests on dead specimens. All the results, particularly the limited absorption of arsenic in the absorption tests, indicated that the isaza were exposed to arsenic before the mortality event. Parallel to this, the geochemical behavior of manganese and arsenic in oxygen-rich conditions (June) and oxygen-poor conditions (December) was investigated to verify the mechanism of exposure. Considerable enrichment of manganese and arsenic in a thin surface layer of sediment was a common feature in all seven stations studied. In the water at the bottom of the lake, a clear increase of arsenite in December was observed, and the manganese level was several hundred times higher in both seasons than the average level of the lake. Although further verification is needed, the data provided here support exposure to manganese and arsenic under hypoxia.