With an autotrophic, N-flee medium, Xanthobacter populations were isolated from the roots of wetland rice grown under field conditions. Xanthobacter populations ranged from 3.2 x 10 4 to 5.1 × 10 5 colony-forming units (cfu) g-~ of root and averaged 47-fold higher on the root or rhizoplane than in the neighbouring nonrhizosphere. Characterization studies indicated dissimilarities in carbon utilization and motility among the isolated Xanthobacter strains and other recognized Xanthobacter species. Under gnotobiotic conditions, the population of one isolate, Xanthobacter sp. JW-KR1, increased from 105 to 107 cfu plant-l 1 d after inoculation when a rice plant was present, but declined to numbers below the limit of detection (<104cfu assembly -l) after 3d in the absence of a plant. Scanning electron microscopy revealed Xanthobacter as pleomorphic forms on the rhizoplane. To assess the effect of Xanthobacter on plant growth, rice plants were grown under greenhouse conditions in plant assemblies containing sand and half-strength Hoagland's nutrient solution with and without nitrogen. Plants were either inoculated with 105 cfu Xanthobacter g-t of sand or left uninoculated. After 40 d, plants without nitrogen showed no significant differences in top or root dry weight, plant height, root length, or number of tillers or leaves, whether the plants were inoculated or uninoculated. However, when nitrogen was added, inoculated plants had a significantly larger top dry weight (15%) and number of leaves (19%) than uninoculated plants. Under conditions of added and no added nitrogen, acetylene reduction assays showed Xanthobacter sp. JW-KR1 produced <0.1 (below detection limit) and 7 nmol C2H 4 plant-~ h-1, respectively. Under the conditions studied, the results suggest that both Xanthobacter and wetland rice derive some benefits from their association.