Purpose : To evaluate potential usefulness of dynamic Gd-DTPA enhanced MRI in the diagnosis of $HgCl_2$ induced acute tubular necrosis of rabbits. Materials and Methods : Sixteen rabbits were used as control group, and 14 rabbits and 12 rabbits were used as acute tubular necrosis groups of 24 hours and 48 hours after $HgCl_2$ injection, respectively. Sequential dynamic MR imagings were acquired using Gd-DTPA(0.25mmol/kg), and time-signal intensity curves were obtained from cortex, outer medulla and inner medulla. Results : In control group, a dark band, which reflects concentrated Gd-DTPA, migrated from cortex to inner medulla of the kidney, and the ratio of the signal intensity of post Gd-DTPA injection to the signal intensity of pre Gd-DTPA injection(RSI) decreased below 1 at 13sec and 26 sec(mean: 17$pm$6.2 sec) in cortex, at 52 sec(mean: 52 sec) in outer medulla, and after 117 sec(mean:112$pm$33.9 sec) in inner medulla of the kidney. In acute tubular necrosis group of 24 hours after $HgCl_2$ injection, the dark band did not appear and signal intensity in cortex and medulla increased diffusely, and RSI increased above 1 in all locations. In acute tubular necrosis group of 48 hours after $HgCl_2$ injection, the dark band appeared only in the cortex and no sign of migration was observed, and RSI is little changed except In cortex at 13 sec(0.76$pm$0.05) and 26 sec(0.86$pm$0.06). There were statistically significant differences in the time-RSI curves among cortex, outer medulla, and inner medulla in study groups, respectively (p<0.0001). Conclusion : Dynamic Gd-DTPA enhanced MRI of the kidney could be utilized to evaluate both renal structure and functional changes.