Quantifying the potential effects of petroleum hydrocarbon (HC) spills on aquatic taxa is inherently difficult. Each spill has a unique HC profile which changes by the hour, limiting the value of toxicity thresholds generated from oil exposure ecotoxicity studies. The derivation of species-specific sensitivity constants (CTLBBs) for aquatic taxa, coupled with the measured or modelled composition of dissolved HCs enables the estimation of acute and or chronic toxicity for any composition over the duration of a spill. This presentation introduces the experimental approaches to generate and validate CTLBBs and highlights opportunities for their application in ecological damage or risk assessments. CTLBBs are derived from a structure-function (toxicity) relationship based on a series of ecotoxicity experiments with individual, representative HCs, which allows toxicity prediction for all other HCs. In a complex oil mixture, the combined effects of all HCs are assumed additive. The Target Lipid Model (TLM, incorporating 79 mostly temperate CTLBBs) is applied to calculate oil concentrations intended to protect marine life from laboratory exposures of a partially weathered condensate. These predictions are then contrasted with empirical results obtained using traditional multi-species ecotoxicity tests in a tropical marine case study. Recent derivation of CTLBBs for tropical marine species allows direct sensitivity comparisons against all species in the TLM database, and enables the toxicity of any oil composition to be modelled. While toxicity values generated by the TLM are employed in oil spill damage and risk assessments internationally, further research is needed to support application to local spill types, conditions and habitats (i.e. coral reefs). Priorities for new research to progress the development, validation and application of HC toxicity modelling are discussed.