In recent years, the inclusion of uncertainty analysis in risk assessment has been much debated. One pertinent issue is the translation of the effects observed with a limited number of test species to a general protection level for most or all species present in the environment. In a number of cases, toxicity data may consist of data from tests employing only a control and one treatment. Given that more species (or processes) have been tested with the same treatment, the treatment can be considered as fixed, and the effect level of the individual species (or processes) can be considered as variable. The distribution of effects can be viewed as a species effect distribution for that treatment. The distribution will represent all organisms and may be used to predict the maximum impact on any fraction of all organisms (e.g., 95% of all species). Hence, it is possible to predict the maximum effect level, with a selected certainty, for a given fraction of all species.