AbstractThe data from a previously reported large series (1067 cases) of insulin tumours of the pancreas have allowed study of the problem of management of so‐called ‚occult’︁ tumours. Of the occult tumours in the whole series, 54 per cent were situated in the head of the pancreas, as opposed to 26.5 per cent similarly located at the initial operation.Analysis of the records of 542 single benign beta‐cell adenomas showed no correlation between the size of the tumour and the duration of symptoms. There were 168 cases of single benign tumours which could not be located at the first operation, but which were cured at a subsequent operation. In 65 per cent of these cases the tumour could not be located at the subsequent operation and was only found after blind resection.These results indicate that the growth of these tumours is not regular and constant. They lend no support to the suggestion that if an insulin tumour cannot be located blind resection should be avoided in the hope that after 1‐2 years, during which symptoms can be controlled by diazoxide, the tumour is likely to he located at a second operation. It is concluded that if the tumour cannot be located the treatment likely to give the best results is progressive blind pancreatic resection from left to right, with careful examination of the excised specimen and frequent estimation of the blood sugar co