To establish a new method for evaluating renal function using MRI, we performed dynamic Gd-DTPA enhanced MR imaging at 1.5 T in 68 subjects (71 examinations), including 23 normal volunteers, 19 patients with chronic renal failure, 22 with hydronephrosis, and 7 with renal vascular disorders. Thirty GRASS images [35/9/20 degrees /1 (TR/TE/Flip Angle/NEX)] were obtained after a bolus injection of Gd-DTPA (0.2 mmol/Kg) in each case, and were evaluated in the following respects. (1) Whether or not the initial signal increase and/or the following signal drop in cortex was observed. (2) Whether or not the signal drop in medulla and/or calyces was observed. (3) Time between initial signal increase in cortex and signal drop in medulla (Ta). (4) Time between signal drop in medulla and calyces (Tb). (5) Time between signal increase in cortex and signal drop in calyces (Tc). (6) Maximum ratio of signal intensity between cortex and medulla (Max. C/M). In normal subjects, the initial signal increase and the following signal drop in cortex, and the signal drop in medulla and calyces were observed. In patients with abnormal renal function, the characteristic signal changes described above were not seen clearly or disappeared in many cases. In normal subjects, Ta, Tb, Tc, and Max. C/M were 41.9 +/- 6.9 sec, 53.2 +/- 10.6 sec, 95.0 +/- 9.5 sec, and 1.10 +/- 0.05, respectively. In the patients, time parameters were significantly larger than those of normal subjects and C/M ratios were lower. This study suggested that dynamic Gd-DTPA renal MR imaging can serve as a new method for the evaluation of renal function.