The toxicity of volatile hydrocarbons to terrestrial organisms is currently being investigated. It was found that toxicity test methods required procedural modification during test soil preparation to minimize the loss of the test substance due to volatilization. A series of tests was initiated with an artificial and field-collected reference soil to investigate alternative test-soil, preparation, methodologies for evaluating motor gasoline (mogas) as it predominantly contains highly volatile low-end carbon components (⩽C5–C13). The proposed approaches to minimize volatilization losses included the application of mogas at concentrations sufficiently high to accommodate the percentage lost during test soil preparation; the modification of test soil preparation methods; and the modification of the nature of test units. The analyses of benzene, toluene, ethylbenzene, xylene (BTEX) compounds, total purgeable hydrocarbons (TPuH), total extractable hydrocarbons (TEH), and total petroleum hydrocarbons (TPH) from soil samples quantified the losses of these mogas constituents at each stage of preparation. The constituents of the mogas were found to volatilize at different rates and the losses were to some extent, concentration dependent. Amending test soils at higher concentrations than required for an adverse effect compensated for the rate of the losses due to volatilization. Significant amounts of the mogas were lost regardless of the methods used to prepare the test soils. Alternative “closed” (minimized air exchange) test units were considered useful and acceptable, as they optimized exposure of the test organism to mogas and did not compromise their survival. The test methods developed to assess the toxicity of mogas in terrestrial organisms may be equally applicable to the toxicity testing of other highly volatile substances.