The enhancing removal of kerosene (jet fuel) from contaminated soil during bioventing resulting from biodegradation was compared to the physical removal by evaporation only on bench-scale columns at the controlled temperature of 20 degrees C (+/- 2.5 degrees C). Carbon dioxide-free air and nitrogen were used as flushing gases, at the constant continuous flow rate of 1 dm(3)/h. Kerosene concentrations in soil up to 35000 mg/kg were not toxic for indigenous microbial population. Much slower kerosene biodegradation rates observed for soil from a contaminated site, as compared to soil artificially contaminated with kerosene, were the result of a lower bioavailability of "aged" kerosene, and the presence of compounds that might be persistent or toxic to kerosene-specific degraders. The inhibitory effect of toluene to indigenous microorganisms was found at above 75% of the toluene saturation concentrations in the gas phase. After 29 d, the overall bioventing efficiency was 17-23%, depending on whether CO2 production or O-2 uptake was used for calculations, as compared to the removal of 10% when biodegradation was excluded. The increase in efficiency by 50-100% owing to biodegradation would be more spectacular at lower kerosene concentrations during the "tailing" phase, with diffusion-limited desorption, and much lower evaporation of less volatile constituents. Limitation of bioventing as a result of low bioavailability related to intraparticle sorption of residual contamination is discussed.