Environmental management decisions are increasingly based on the results of ecological risk assessments that require data and models to establish the likelihood of adverse ecological effects that may occur or are occurring as a result of exposure to one or more stressors .. Research to support ecological risk assessments for chemical stressors must include a predictive capability.. The great number of chemicals, species and habitats (exposure conditions) associated with prospective and retrospective management decisions precludes the use of empirical testing as the sole means to fill data gaps .. Therefore, strategic laboratory and field testing must be undertaken in the context of a modeling strategy that is based on ecologically-relevant toxicological endpoints and mechanistically-sound toxicological hypotheses .. The development of quantitative structure activity relationships (Q§ARs) to predict the toxicity of untested chemicals has evolved from a chemical-class perspective to one that is more consistent with assumptions regarding modes of toxic action. To improve the means of extrapolating adverse effects across species and exposure regimes, physiologically-based toxicokinetic and toxicodynamic models must be linked with a greater array of mechanistically-based Q§AR techniques.