Although environmental contaminants are usually encountered as nonequitoxic mixtures, most studies have investigated the toxicity of equitoxic mixtures. In the present study, a method for prediction of the toxicity of nonequitoxic mixtures was developed using the similarity parameter (λ). The joint effect of multiple contaminants at the median inhibition concentration in equitoxic ([Formula: see text]) and nonequitoxic ([Formula: see text]) binary, ternary, and quaternary mixtures was investigated using Vibrio fischeri. The observed results indicate that the concentration ratios of individual chemicals in the mixtures influenced the joint effects, and that λ could be employed to evaluate the relation between [Formula: see text] and [Formula: see text]. Prediction models for the joint effects of nonequitoxic ([Formula: see text]) mixtures were derived from a combination of [Formula: see text] and λ. The predictive capabilities of these models were validated by comparing the predicted data with the observed data for binary, ternary, and quaternary mixtures. The prediction models have promising applications in controlling environmental pollution, evaluating drug interactions, and optimizing combinations of pesticides used in agriculture.