The effect of major ions (Ca, Mg, Na, and K) and pH on Cu toxicity (LC50) to Hyalella azteca was determined in 1 week exposures. The simplest equation for describing Cu toxicity is a linear relationship between the total dissolved Cu LC50 and Ca and Na in water, ignoring pH. This equation would be useful in tier one of a two-tiered approach; if the measured dissolved Cu exceeds the value predicted from the equation, the sample should either be tested for toxicity, or a more detailed chemical speciation analysis can be conducted. The data were not consistent with a single-binding-site biotic ligand model, assuming that toxicity was due to the free Cu ion alone. However, toxicity could be predicted using a two-binding-site model. This requires separate coefficients to account for the effects of Ca and Na at low and high pH values (6.5-8.4), corresponding to the different binding sites (Mg and K did not affect toxicity). The single-binding-site BLM does not allow for this. Toxicity of Cu hydroxide or carbonate complexes does not need to be invoked, but cannot be excluded, and several models invoking the toxicity of these complexes can also explain the data. The free ion LC50 is strongly dependent on pH, but the LC50 for total dissolved Cu is almost pH independent. The effects of Ca and Na on the free ion LC50 are very different at high and low pH (contrary to single-site biotic ligand model predictions), but similar for total dissolved Cu. Published data suggest that the same model, with different coefficients, can also be applied to Daphnia and fish. A more critical evaluation of the effects of cations at both low and high pH for organisms other than Hyalella is needed to determine if the BLM needs to be adjusted to incorporate more than one binding site for other species as well. Hydrogen ions reduce the toxicity of free Cu ions to Hyalella, but Cu also reduces the toxicity of hydrogen ions. A mixture model accounting for the joint toxicity of Cu and pH, as well as their mutual antagonistic effects, is presented.