Environmental risk assessment of chemicals is mostly based on the results of standardized toxicity tests. To obtain environmental quality criteria, extrapolation factors are used that depend on the amount and quality of available data. These extrapolation factors do not, however, take into account the mode of action of the compound tested or the life history of the test organism. In this study, we analyzed the variability in acute-to-chronic ratios (ACRs) for various chemicals in relation to their mode of action. Chemicals were classified as nonpolar narcotics, polar narcotics, specifically acting compounds, and heavy metals. As an acute endpoint, the LC50 was used; as a chronic endpoint, the lowest test concentration at which the natural rate of population increase (r) is affected, or LOEC(r), was used. Data were derived from the on-line literature. Nonpolar narcotic chemicals demonstrate the smallest variation in ACRs, and acute tests can be used to derive chronic endpoints for this class. For the other classes, the variation in ACRs is larger. Fish species especially show a relatively large ACR. For heavy metals, differences in the mode of action may play an important role in explaining differences in ACRs. For the other three classes, however, it is less reliable to predict chronic toxicity using the results of acute tests. In general, differences in species sensitivity rather than in mode of action for the chemical seem to determine differences in ACRs.