On military training ranges, low-order, incomplete detonations deposit RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine) into surface soils. In this study, we evaluated RDX biodegradation in surface soils obtained from a military training range in Alaska. Two factors were compared: (i) soil water potential during the incubations; and (ii) the use of acetonitrile (ACN) as an RDX carrier to spike samples. Organic solvents have been used in laboratory studies to dissolve slightly water-soluble contaminants before addition to soil. We added ACN to obtain final soil ACN concentrations of 0 mg kg(-1) (0%), 1000 mg kg(-1) (0.1%) and 10 000 mg kg(-1) (1%). We then compared RDX attenuation in the soil under saturated and unsaturated conditions. RDX fell below the limit of detection within 3 wk of study initiation under the saturated condition. A maximum degradation rate of 0.15 mg RDX L(-1) d(-1) was measured. Under the unsaturated condition, 42% of the original RDX was still present at study termination (5 wk). The addition of acetonitrile at 0.1 or 1.0% had no affect on RDX loss in the saturated soil. In the unsaturated soil, however, ACN at 1.0% inhibited RDX loss by as much as 25%. These findings indicate that soil water potential and carrier solvent concentrations can impact the rate and extent to which RDX is attenuated in a surface soil.