There is a concern that concentration-based targets for soil clean-up are arbitrary, and do not necessarily relate to the environmental threat posed by the residual contamination. The development of sensitive, reliable, and ecologically relevant biological tests for oil-polluted soils would address these problems, and form a valuable complement to chemical analysis. Three biological test systems were chosen for examining the impact of oil contamination on soil microbes: (1) lux-marked bacterial biosensors, (2) Biological carbon source utilization profiles, and (3) phospholipid fatty acid (PLFA) community profiles. This allowed the testing of effects at three different levels of ecological complexity: respectively, single species; culturable bacteria; and eubacteria and eukaryotes. The use of specific bioluminescent biosensors (i.e. with lux genes fused to promoters from hydrocarbon degradation pathways) allowed the rapid detection of different hydrocarbon classes. The bacterial biosensors were optimized for the assessment of hydrocarbon compounds. Development of quantitative structure-activity relationships (QSARs) increased understanding of their responses to hydrocarbons and other petroleum-derived compounds. The three biological methods were then used to monitor the nutrient-assisted bioremediation of oil-spiked soils in two separate microcosm experiments: (1) remediation of four crude oils and one refined oil compared using bacterial biosensors, and (2) effects of diesel on three different soil types, and effects of three refined oils on one soil type, assessed using Biolog and PLFA profiling. All three biological test methods were sensitive to the oil contamination levels applied in the microcosm experiments. Individual species (biosensor tests) gave different responses to different oils; however, the community-level responses showed no differences. The microbial communities of the three different soil types could still be distinguished after 14-15 weeks of hydrocarbon contamination. The community response of the contaminated soils had changed from that of the control soils after four weeks of bioremediation, and did not return in similarity to the control over the course of the experiment.