Since the discovery of the hormone (Bonting and Best, 1922), extensive research on insulin has been going on. Insulin is important not only in relation to carbohydrate metabolism, and thus in relation to diabetes mellitus, but also as one of the important hormones In protein and fat metabolism. There is much evidence showing an impairment of insulin action in stress, whether the stress is physical or psychological. The way in which insulin resistance develops and its effects on metabolism may vary in different stress situations. Th is work investigates three forms of physical stress and attempts to show the role of insulin In each of these situations. The metabolic and clinical implications o f this problem are discussed. First, the effects of surgery were studied in eleven patients. Their mean age was 52 years, mean weight was 68 kg, and mean % ideal body weight was 94%. A nitrogen balance study showed that the negative balance after surgery coincided with elevated levels of plasma glucagon, non-esterified fatty acids, branched-chain amino acids, urinary free cortisol, urinary 1 7-OH-corticosteroids and with a decrease of total plasma amino acids. A temporary insulin resistance occurred in post-surgical patients, shown by hyperglycaemia and hyperinsulinaemia during a two hour glucose infusion. The clinical significance of hyperglycaemia is discussed. The second study was of twenty nine obese patients (fasting blood glucose ^6.11 m m ol/l). Their mean age was 40 years, mean weight was 109 kg , and mean % ideal body weight was 166% (from 115% to 233% ). These patients showed on average an impaired response to the oral glucose tolerance test (iv GTT) , intravenous glucose tolerance lest (iv GTT) and intravenous glucose insulin tolerance lest (iv GTT) for insulin sensitivity. Th is impairment was related to hyperinsulinoemia which followed glucose administration (oral or iv ). These obese patients seemed to fa ll into two groups: those with % ideal body weight K 160% showed impaired tolerance to glucose but relatively normal plasmo insulin responses; those >160% showed marked hyperlnsulinaemla. It is suggested that these responses represent those o f 'active' and 'passive' obesity, but that the former may include a pre-clinical stage when insulin sensitivity Is very high. Thirdly, twenty three geriatric patients (fasting blood glucose ^ 6.11 m mol/1) -were studied. Their mean age was 79 years and mean weight was 56 kg. These patients also showed an Impairment in the oral GTT , Iv GTT and iv GTT . The impairment was greater than that found in obese patients. Insulin response to glucose administration (oral or iv) was sluggish, but the actual levels of insulin were not significantly lower than those found in young normal subjects (except for the peak value during iv GTT). The major cause of impaired glucose tolerance was diminished insulin sensitivity, either in the peripheral tissue, or, more probably in the liver, resulting In relative in ability to switch off glucose output. An Intravenous alanine tolerance test was carried out in eight elderly subjects (mean age was 78 years, mean weight was 52 kg), to assess gluconeogenic capacity of the liver, and again indicated the relative inability o f endogenous insulin to suppress glucose production.