The study aimed to assess vascular reactivity to noradrenaline with and without neuropeptide Y in diabetic rats, and to determine whether any abnormality could be attributed to insulin deficiency or to hyperglycaemia per se. The authors compared non-diabetic rats (n = 9) and rats with streptozotocin-induced diabetes that were either untreated (n = 10), or treated with insulin (n = 9) or food restriction (n = 8) to restore near-normoglycaemia. After 4 weeks of diabetes, contractile responses to noradrenaline (0.24-48 mumol L-1), without and with neuropeptide Y (0.1 mumol L-1), were assessed using an isometric myograph in two mesenteric arteries from each rat. Vessels from untreated diabetic rats were significantly more reactive to noradrenaline than the control vessels when tested without (P < 0.0001) but not with (P = NS) neuropeptide Y. Diabetic rats rendered nearly normoglycaemic through food restriction showed dose-response curves that were very similar to the untreated diabetic group (P = NS). By contrast, insulin-treated diabetic vessels showed reduced sensitivity to noradrenaline, with and without neuropeptide Y, compared with both the diet-restricted and untreated vessels (both P < 0.0001). The authors conclude that vascular sensitivity to noradrenaline, without or with neuropeptide Y, is reduced over a wide dose range in vessels taken from rats treated in vivo with insulin; furthermore, vessels taken from diabetic rats not treated with insulin (hypoinsulinaemic) tended to be more reactive than either control vessels or those taken from the insulin-treated rats. The latter group of rats were probably hyperinsulinaemic for much of the time; the results may therefore support the hypothesis that insulin acts as a vasodilator.