The species sensitivity distribution, a technique currently used to derive water-quality standards of chemicals, is associated with a set of inadequately tested assumptions. One of these assumptions is that ecosystem structure is as or more sensitive than ecosystem function, i.e., that structure is the target of concern. In this paper, we tested this assumption for a simple freshwater ecosystem exposed to different toxicants. Using an ecosystem model, we calculated no observed effect concentrations (NOECs) for ecosystem structure (ecosystem structure-NOECs) and function (ecosystem function-NOECs) for each of 1000 hypothetical toxicants. For 979 of these toxicants, the ecosystem structure-NOEC was lower than or equal to the ecosystem function-NOEC, indicating that the tested assumption can be considered valid. For 239 of these 979 toxicants, both NOECs were equal. For half of the 1000 toxicants, the structure of lower trophic levels (i.e., phytoplankton) appears to be more sensitive than the structure of higher trophic levels (i.e., fish). As such, ecosystem structure-NOECs are primarily determined by the sensitivity of the structure of lower trophic levels. In contrast, ecosystem functions associated with higher trophic levels (e.g., total ingestion by fish) are more sensitive than functions associated with lower trophic levels (e.g., total photosynthesis by phytoplankton) for 749 toxicants.