On-site methods using the seed germination test to evaluate the phytotoxicity of hazardous waste site soils are advantageous because samples are not removed from the site and time is saved in soil handling and preparation. Seedlings may be collected at the end of the test and taken to the laboratory for further analyses. We have used the Magenta™ GA-7 test vessels to conduct on-site the germination of cucumber, Cucumis sativus L. var. Hybrid Calypso, and wheat, Triticum aestivum L. var. Stephens. Site soil was sampled using a premeasured cup that held a volume nominally equivalent to 100 g of soil. Twenty pregraded seeds were planted in the soil in each test container and covered with 16-mesh silica sand. Site soil was irrigated with water to approximately 45% moisture on a dry weight basis, covered, and left under a shade at the test site for five days. After five days, germinated seeds were counted and percent germination values were calculated for each seed species. Controls for the test were performed on-site, using 16 and 20-mesh silica sand. Test performance was evaluated against the companion laboratory seed germination test. Calculated percent germination values in the cucumber seeds indicated statistically significant (α = 0.05) but biologically reasonable differences between the laboratory test and the field test. Cucumber seeds showed a greater difference, while the wheat seeds did not show any significant difference between the tests.