Plasma ACTH levels in response to metyrapone and insulin hypoglycaemia were compared in subjects with normal pituitary-adrenal function. After a single dose of 2 g of metyrapone given with a snack at midnight, the ACTH level was 468 ng/l +/- 66 )SEM) at 07.30 h the next morning (mean increment approximately nine fold over normal morning values). After insulin-hypoglycaemia the peak ACTH level was 369 ng/l +/- 31 (SEM). Peak ACTH levels greater than 200 ng/l were achieved in twenty of twenty-one (95%) subjects after metyrapone and twenty of twenty-four (83%) after insulin. No major side effects were noted after metyrapone. It is concluded that the short single-dose metyrapone test produces at least as strong and consistent a stimulus to ACTH release as the standard insulin-hypoglycaemia test in normal subjects. A direct assay of ACTH avoids misinterpretations which are inherent in a judgement based on compound S increase only. The short test has significant practical advantages over the classical metyrapone test, and provides a convenient and sensitive method of assessing the negative feedback ACTH control mechanism. It may be particularly useful in detecting minor degrees of pituitary suppression. The value of this test in clinical practice for the investigation of patients with hypothalamic-pituitary diseases in comparison to the classical tests of ACTH stimulation has yet to be demonstrated.