A case of insulinoma whose serum showed 125I-insulin binding presumably due to insulin autoantibody is reported. A 60-year old male was admitted because of morning disorientation and convulsion attacks, and was proven to have hypoglycemia and high serum insulin values at fasting. There was no history suggesting the injection of insulin in the past. Serum IRI values increased moderately after various stimulation tests. Binding of 125I-insulin to serum was demonstrated by polyethylene glycol method, gel filtration and hydrodynamic flow on a filter paper. This binding was inhibited by the addition of enough amount of unlabeled insulin. The precipitation experiments with anti-human immunoglobulin fractions suggested that insulin antibody is probably responsible;or this binding. Dilution test, recovery test and extraction of the serum suggested that the interference with IRI assay by this binding was negligible. By laparotomy, an adenoma, 1.3 cm in diameter was found in the tail of the pancreas. The tumor was rich in capillaries and contained a few scattered cells with the staining characteristics of B-cells by aldehyde-fuchsin and immunofluorescent methods. Electron microscopy revealed that most cells were agranular but a few cells contained secretory granules resembling those in normal B-cells. Insulin content of the tumor was2.6U/g and lower than that in the adjacent normal pancreas tissue (4.6U/g). Hypoglycemic symptoms and 125I-insulin binding of serum disappeared after the operation. The coexistence of insulinoma and “insulin autoantibody” such as found in this patient casts a new problem on the diagnosis of insulin autoimmune syndrome, although the causal relationships of the tumor and “antibody” remain unsolved.