Testing for the ecological impacts of toxins can occur in the lab or in the field. Statistical design concepts can provide direction for the spacing and allocation of experimental organisms for optimal estimation of system characteristics. Inference tools from statistics can be used to provide methods for customizing analyses for special experiments or for suggesting better analysis options for standard experiments. This talk will present a series of case studies describing optimal design for estimating uptake and elimination rate constants, developing a more powerful test for step-stress experiments, avoiding overly conservative inference in simple two group comparisons, and incorporating prior information in potency estimation.