The leaching behaviour of quinclorac was elucidated using soil columns. On top of each glass column packed with a rice paddy soil up to the 30 cm height were applied three different treatments of [$^{14}C$]quinclorac: quincloiac only (T-1), quinclorac adsorbed onto active carbon (T-2), and quinclorac adsorbed onto a mixture of active carbon and $Ca(OH)_{2}$ (T-3). Half of the columns were planted with rice plants for 17 weeks and half of them unplanted for comparison. Average amounts of $^{14}C$-activity percolated from tile soil columns without rice plants in T-1, T-2, and T-3 were 81.1%, 27.8% and 48.0%, respectively, of tile originally applied $^{14}C$, whereas those with rice plants grown were 36.8%, 9.6% and 11.0%, respectively, indicating that the leaching of [$^{14}C$]quinclorac was significantly affected by vegetation and by treatment with the adsorbents. The bioavailability of the herbicide to rice plants in T-1, T-2, and T-3 were 13.6%, 11.0% and 13.9%, respectively. The residue levels of quinclorac in the edible part of rice grains would be far less than the maximum residue limit (MRL, 0.5 ppm). After the leaching, the amounts of $^{14}C$ remaining in soil in with rice planting T-1, T-2, and T-3 were 36.3%, 73.7%, and 61.8%, whereas those without rice planting were 19.7%, 71.1%, and 52.3%, respectively. The balance sheets indicate that [$^{14}C$]quinclorac translocated to rice shoots would be lost by volatilization and/or in other ways in T-1 and T-3. The $^{14}C$-activity partitioned into the aqueous phase of the leachates collected from all treatments was less than 7% of the total, but it increased gradually with time in the case of rice growing, suggesting tile formation of some polar degradation products.