Extrapolating toxicant effects with a fixed application factor (AF) approach or one of the species sensitivity distribution (SSD) models presumes that toxicant effects on single, individual-level endpoints reflect effects at the ecosystem level. Measured effect concentrations on plankton from multispecies field tests using tributyltin (TBT) and linear alkylbenzene sulfonates (LAS) were compared with published laboratory single-species test results and measured in situ concentrations. Extrapolation methods were evaluated by comparing predicted no-effect concentrations (PNECs), calculated by AF and SSD models with NOECs and E(L)C₅₀s obtained from field studies. Overall, structural parameters were more sensitive than functional ones. Measured effect concentrations covered approximately the same range between laboratory and field experiments. Both SSD and AF approaches provide PNECs that appear to be protective for ecosystems. The AF approach is simpler to apply than the SSD models and results in PNECs that are no less conservative. Calculated PNEC values and the lowest field effect concentrations were lower than measured environmental concentrations for both substances, indicating that they may pose a risk to marine ecosystems.