The purpose of the study was to evaluate the potential of dynamic Gd-DTPA-enhanced MR imaging in assessing of the function of transplanted kidneys. Dynamic MR imaging was performed using the GRASS pulse sequence in 21 subjects (56 examinations), and in each examination 30 images were obtained after a bolus injection of Gd-DTPA. Imaging findings including signal increase and/or signal drop in cortex, signal drop in medulla, and signal drop in calyces were utilized for the evaluation of renal function. Renal function was also evaluated using the following parameters: Ta, time between the initial observation of signal increase in cortex and signal drop in medulla. Tb, time between the initial observation of signal drop in medulla and in calyces; Tc, Ta + Tb, and Max. C/M, maximum ratio of signal intensity between cortex and medulla among 30 images. In the group of patients with normal graft function (Ccr > or = 60ml/min), all imaging findings were observed. In the group of patients with mild graft dysfunction (30ml/min < or = Ccr < 60ml/min), all imaging findings but signal drop in cortex were observed. The time parameters of Ta, Tb, and Tc were significantly larger, and Max. C/M was significantly smaller than those observed in the group with normal graft function. In the group of patients with severe graft dysfunction (Ccr < 30ml/min), almost none of the findings were observed. The investigation in patients in the postoperative state revealed a close correlation of the imaging findings and the values of parameters with changes in graft function over time. The results indicated the usefulness of the method in semiquantitative evaluation of graft function, including that of patients in postoperative state. This study suggests that dynamic Gd-DTPA-enhanced MR imaging could be a valuable method for the management of transplanted kidneys.