The uptake of hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) from soil by the earthworm Eisenia andrei was examined by using the equilibrium partitioning (EqP) theory and a three-compartment model including soil (S), interstitial water (IW), and earthworms (E). The RDX concentrations were measured using U.S. Environmental Protection Agency (U.S. EPA) Method 8330A and high-performance liquid chromatography (HPLC). The S-IW studies were conducted using four natural soils with contrasting physicochemical properties that were hypothesized to affect the bioavailability of RDX. Each soil was amended with nominal RDX concentrations ranging from 1 to 10,000 mg/kg. The HPLC analysis showed that the IW extracted from soil was saturated with RDX at 80 mg/kg or greater soil concentrations. The calculated S-IW coefficient (K(p)) values for RDX ranged from 0.4 to 1.8 ml/g soil, depending on the soil type, and were influenced by the organic matter content. In the IW-E studies, earthworms were exposed to nonlethal RDX concentrations in aqueous media. The uptake of RDX by the earthworms correlated well (r(2) = 0.99) with the dissolved RDX concentrations. For the E-S studies, earthworms were exposed to RDX-amended soils used in the S-IW studies. The bioconcentration factors (BCF; ratios of E-to-IW RDX concentrations) were relatively constant ( approximately 5) up to 80 mg/kg soil RDX concentrations, which encompass the RDX saturation limit in the interstitial water of the tested soils. At this concentration range, the RDX uptake from interstitial water was likely dominated by passive diffusion and could be used as an indicator of bioavailability. Other mechanisms may be involved at greater RDX soil concentrations.