In order to evaluate the safety and pharmacokinetics of gadolinium contrast agents in patients with hemodialysis, in vitro and clinical studies were performed. Gd-DTPA and Gd-HP-DO3A solutions were dialysed by standard methods. One classic type (cellulose acetate) and two high-performance types (PMMA and polyacrylonitrile) were used. Eleven dialysis patients underwent normal dose contrast MRI followed by usual dose hemodialysis (HD). PMMA dialysis membranes were used for 8 patients and cellulose acetate membranes for 3 patients. Serum levels of gadolinium, BUN and creatinine were analyzed before and after the 1st HD and after the 2nd, 3rd and 4th HD. Side effects and edema were carefully observed. In vitro studies showed that 11.1 hours of dialysis would be necessary to remove 97% of the injected dose of contrast agent by using a first-order kinetic model of dialysis time. When dialysis membranes and contrast agents were compared, statistically significant differences were noted between dialysis membranes and all contrast agents. In the clinical study, 79.1% of the contrast agent was dialyzed after the 1st dialysis and 99.6% after the 4th dialysis. These results showed that all contrast agents and both dialysis membranes were suitable. Neither changes in laboratory parameters nor side effects were observed. The present study suggests, consequently, there are no contraindications when using the ordinary dose of contrast agent even in patients with dialysis.