The biotic ligand model (BLM) for the acute toxicity of cationic metals to aquatic organisms incorporates the toxicity-modifying effects of dissolved organic matter (DOM), but the default parameterization (i.e., assuming 10% of DOM is humic acid) does not differentiate DOM from different sources. We exposed a cladoceran (Ceriodaphnia dubia) to Ag in the presence of DOM from filtered YCT (standard yeast-Cerophyll(R)-trout chow food recommended by the U.S. Environmental Protection Agency [EPA] for cladocerans), from the Suwannee River (GA, USA; relatively little anthropogenic input), and from the Desjardins Canal in Hamilton (ON, Canada; receives treated municipal wastewater effluent). In all three treatments, the dissolved organic carbon (DOC) concentration was 2 mg/L (the concentration following addition of YCT slurry at the U.S. EPA-recommended volume ratio). The average 48-h median effects concentration (EC50) ratios for dissolved Ag in the presence and absence of DOM [i.e., (EC50 with DOM)/(EC50 without DOM)] were as follows: Suwannee River, 1.6; Desjardins Canal, 2.2; and YCT filtrate, 26.8. Therefore, YCT filtrate provided much more protection against Ag toxicity than that provided by DOM from the surface waters. The major spectral characteristic that differentiated YCT filtrate from the other two types of DOM was a strong tryptophan peak in the excitation- emission matrix for YCT. These results have important implications for interpreting Ag toxicity tests in which organisms are fed YCT, and they suggest BLM-calculated toxicity predictions might be improved by incorporating specific chemical constituents or surrogate indices of DOM. Another component of the protective effect against Ag toxicity, however, might be that the dissolved fraction of YCT served as an energy and/or nutrient source for C. dubia.