A bioassay for rhizosphere-applied phytotoxicants was developed and evaluated with a broad range of chemicals. Test substances were applied to the rhizosphere of whole, intact bush bean plants (Phaseolus vulgaris L. cv. Bush Blue Lake 290) grown in a solid support medium and the resultant ethylene production was measured to detect the presence of phytotoxic materials. The beans were encapsulated in plastic bags for 2 hrs following treatment and then incubated for 24 hrs in the dark. Ethylene and ethane accumulating within the bags were quantified via gas-solid chromatography. The application of various concentration of inorganic and organic chemicals induced various responses. No single equation adequately described the dose-response curves; therefore, a critical value (a statistically significant increase in stress ethylene production) was computed for each test substance. A phytotoxic-response threshold for each test substance was defined as the test substance concentration that caused ethylene production to exceed its respective critical-value concentration. Based on threshold concentrations determined by analysis of stress ethylene production, the relative phytotoxicity rankings of the inorganic test substances were: CdCl2 > CuCl2 > Pb(C2H302)2 > LiCl, while those of the organic test substances were 2,4-D esters < paraquat dicloride < 1-butanol < 2-propanone < 2-propanol. Both stress ethylene and ethane production were nonresponsive to hydrogen-ion concentration (measured before application) over a broad range.