We encountered a patient with diabetes mellitus due to the 3243 mitochondrial tRNA mutation(DM-Mt3243), who developed insulin edema and hepatic dysfunction after starting insulin. Such a rare phenomenon was unlikely to be a fortuitous coincidence in mitochondrial diabetes, as none in 197 non-mutant NIDDM patients had same episode. Moreover, similar leg edema was noticed in another DM-Mt3243 patient, and other two DM-Mt3243 patients had leg edema which responded to coenzyme Q10. These observations suggest further a role of mitochondrial function on leg edema. The mechanism of his insulin edema may involve vasomotor changes induced by the rapidly glycemic control, because our case of insulin edema had a prominent increase of strong succinate dehydrogenase reactive vessels. Alternatively, myocardial dysfunction might have produced leg edema and hepatic dysfunction, because he had subclinical myocardial dysfunction, judged by imaging with beta-methyl-p-(123I)-iodophenyl-pentadecanoic acid. The third explanation is that a rapid improvement of glycemic control might have induced hepatic reoxygenation and the production of reactive oxygen species in the liver that contributed to cell damage. Thus, although we cannot draw definite conclusion, our experiences here suggest that mitochondrial dysfunction is important in the etiology of insulin edema.