Traditionally, soil extraction techniques have been concerned with the determination of ``total'' organic contaminant concentrations, following an ``exhaustive'' extraction. However, in light of the increasing body of knowledge relating to organic contaminant availability and aging, such methods have little relevance to the amount of contaminant that may pose an ecological risk i.e., the ``bioavailable'' portion. Less exhaustive techniques have therefore been the subject of more recent approaches in the hope that they may access the ``labile'' or bioavailable pool. The use of an aqueous-based extraction technique utilizing hydroxypropyl-beta-cyclodextrin (HPCD) is presented here for the extraction of PAHs from soil. The optimization of the method is described in terms of HPCD concentration, extraction time, and solution buffering. The procedure is then tested and validated for a range of C-14-labeled PAHs (phenanthrene, pyrene, and benzo[a]pyrene) added at a range of concentrations to a range of soil types. The amounts of soil-associated phenanthrene mineralized by catabolically active microorganisms were correlated with total residual phenanthrene concentrations (r(2) = 0.889; slope of best fit line = 0.763; intercept = -5.662; n = 24), dichloromethane (DCM)-extractable phenanthrene concentrations (r2 = 0.986; slope of best fit line = 0.648; intercept = 0.340; n = 24), butan-1-ol (BuOH)-extractable phenanthrene concentrations (r(2) = 0.957; slope of best fit line = 0.614; intercept = 0.544; n = 24), and HPCD-extractable phenanthrene concentrations (r(2) = 0.964; slope of best fit line = 0.997; intercept = 0.162; n = 24). Th us, in this study, the microbially bioavailable concentrations of soil-associated phenanthrene were best predicted using the optimized HPCD extraction technique. In contrast, the DCM Soxhlet extraction and the BuOH shake extraction both overestimated phenanthrene bioavailability by, on average, >60%.}