In an effort to determine the significance of experimental design on the results of laboratory sediment toxicity studies with amphibians (Xenopus laevis), two different sample preparations were evaluated from three different contaminated waste sites. Whole sediment and aqueous sediment extracts from each site were evaluated. Site 1 soil was characterized as loamy with a relatively high total organic carbon (TOC),moisture fraction (MF), and sulfide content; and contaminated with organochlorine pesticides. Site 2 soil was characterized as silty/clay with low/moderate TOC, MF, and sulfide; and contaminated with polynuclear aromatic hydrocarbons (PAHs) and pentachlorophenol. Site 3 soil samples consisted of two separate subsamples, the first characterized as loamy with a relatively high TOC, MF, and sulfide content, and the second as a mixture of silty/clay and sand with relatively low TOC, MF, and sulfide content. Both sub-site samples were contaminated with heavy metals, including copper, lead, and zinc. FETAX (Frog Embryo Teratogenesis Assay — Xenopus) testing of Site 1 samples indicated that substantially greater levels of developmental toxicity were induced by the aqueous extracts than the whole bulk soil. Tests with Site 2 samples suggested that both of the preparations were capable of inducing comparable rates of developmental toxicity. Tests with subsample a of Site 3 indicated that the aqueous extract of the sample induced greater levels of developmental toxicity than the whole soil. Toxicity tests with subsample b produced variable results that seemed to suggest that the extracts induced greater levels of toxicity than the whole bulk preparations. However, the differences in toxicities noted between that two preparations were not as dramatic as observed in the Site 3a subsample tests. Results from these studies suggested the importance of experimental design in evaluating potential ecological hazards of contaminated sediments or soils, particularly to amphibian species.