The potential bioavailability of phenanthrene aged in soil was determined by using a self-dying reporter bacterium, and the results were compared to two physicochemical measures, Tenax TA (R) bead-assisted desorption, and hydroxypropyl-beta-cyclodextrin (HPCD) extraction. The reporter bacterium, capable of degrading phenanthrene as a sole carbon and energy source, was genetically reconstructed to die when it degrades phenanthrene. Therefore, population change of the reporter cells can be viewed as the quantification of bioavailable phenanthrene. When Ottawa sand was used as an aging matrix, the amounts of bioavailable phenanthrene (i.e. little gradual decrease) were similar, regardless of aging time, and consistent between the reporter bacterium and the two physicochemical measures. However, decrease in bioavailable phenanthrene with aging was readily evident in sandy loam with organic matter of 11.5%, with all three measures. More importantly, when the reporter bacterium was used, a rapid and significant decrease in the bioavailable fraction from 1.00 to 0.0431 was observed. The extent of decrease in bioavailable fraction was less than 40% in the two physicochemical measures, but was nearly 100% in the reporter bacterium, during the first 3 months of aging. Our results suggest that the phenanthrene fraction available to bacterial degradation, and probably the fraction that really manifests toxicity, may be much smaller than the fractions predicted with the physicochemical measures. (C) 2013 Elsevier B.V. All rights reserved.