Currently, when conducting a quantitative health risk assessment, the selection of toxicological reference values (TRVs) relies little on an understanding of the methods used to develop them. As a result, risk assessors tend to choose the value that provides the greatest protection for human health. The purpose of this article is to describe the various methods used to derive TRVs, analyze their key points, and discuss their implications for potential changes in the criteria for selecting TRVs. TRVs are developed for two types of effects: those that occur above a dose threshold (primarily non-carcinogenic effects) and those that occur regardless of dose (carcinogenic effects). Overall, the development process consists of three parts: the selection of a critical effect; the determination of a critical dose; extrapolation to dose levels lower than those observed in the studies serving as the basis for deriving the TRVs. For this final step, the determination of TRVs for dose-threshold effects results from the application of uncertainty factors that allow, among other things, for the consideration of inter- and intra-species sensitivities; TRVs for effects without a dose threshold result from a linear mathematical model that generally passes through the origin. The key factors in this derivation that influence the numerical value of the TRV are, on the one hand, the experimental data used to determine the critical dose and effect, and, on the other hand, the biological, pharmacokinetic, and pharmacodynamic data. The examples of styrene and ethylene glycol monobutyl ether (EGMBE) are used to illustrate this influence. Analysis of current methods for deriving TRVs reveals the following trends: the questioning of the formal separation of derivation methods for the two types of effects, particularly with the development of the benchmark dose; the importance of incorporating biological data to reduce uncertainties in the animal-to-human extrapolation; and the development of uncertainty characterization using probabilistic approaches. These developments call for the development of high-level expertise in toxicology and for the establishment of procedures in France to structure the development of TRVs.