Composting is being explored as a means to remediate 2,4,6-trinitrotoluene (TNT) contaminated soils. This process appears to modify TNT and to bind it to organic matter. The health hazards associated with dusts generated from such materials cannot be predicted without knowing if the association between TNT residues and compost particulate is stable in biological systems. To address this question, single doses of [l4C]-TNT, soil spiked with [14C]-TNT, or compost generated with [l4C]-TNT-spiked soils were administered to rats by intratracheal instillation. The appearance of 14C in urine and tissues was taken as an indication of the bioavailability of TNT residues from compost particles. In rats instilled with neat [14C]-TNT, about 35% of the UC dose appeared in urine within 3 d. Thel4C excreted in urine by these rats decreased rapidly thereafter, and was undetectable by 4 wk after treatment. Similar results were obtained with soil-treated rats. In contrast, after treatment with [14C]-TNT-labeled compost, only 2.3% of the total 14C dose appeared in urine during the first 3 d. Low levels of14C continued to be excreted in urine from compost-treated rats for more than 6 mo, and the total amount of HC in urine was comparable to that in TNT-treated animals. Determination of the radiolabel in tissues showed that MC accumulated in the kidneys of rats treated with labeled compost but not in rats treated with [l4C]-TNT or [14C]-TNT-spiked soil. These results indicate that the association between TNT and particulate matter in compost is not stable when introduced into the lungs. Accumulation of 14C in kidneys suggests the presence of a unique TNT residue in compost-treated rats. The rate of excretion and tissue disposition of UC in rats treated with TNT-spiked soil indicate that TNT in soil is freely available in the lungs.