niques. The study population, consisting of normal individuals with a mean age of just under 30 years with 6 % over 50 years of age, provides an appropriate comparative population for studies in patients with head injury and probably subarachnoid haemorrhage, and could be usefully extended to an older population to allow comparison of normal ranges in patients with internal carotid artery stenosis and occlusion. It is notable that only one subject in the study group had insufficient transcranial Doppler signals from the middle cerebral arteries, representing less than 3 %, thus the range up to 20 % mentioned as a limitation of transcranial Doppler does not apply to this age of study population. The reproducibility of CO # reactivity at 1 h represents a key component of this study. The authors correctly identify that prognostic information on patients with cerebrovascular or carotid artery disease could be studied with such a technique and refer to published work relating to carotid artery stenosis and to carotid endarterectomy. The potential applications are of course rather wider, including intensive care unit monitoring of patients with head injury, assessment of subarachnoid haemorrhage, assessment of comparative drug effects such as the cerebrovascular changes induced by potential neuro-protectants, and the cerebrovascular changes caused by antihypertensive medication, thinking for example of the differences in expected effect between vasodilator drugs and angiotensin-converting enzyme inhibitors. It is important in analysing the results of such a paper to consider whether expected associations which proved negative were definitively assessed. The authors conclude that venous return and sex make no difference to testing CO # cerebrovascular reactivity with this technique. Importantly, the results are consistent with previous transcranial Doppler studies of CO # reactivity and with the smaller previous report of variability in patients with carotid artery disease. The lack of any observed difference according to head position was clearly chosen as a study variable because changes in head position are adopted in surgical and intensive care management of patients with problems affecting cerebral blood flow and cerebrovascular reactivity. Some caution needs to be applied in translating the results from the healthy normal population here to the diseased population where head position is chosen on the basis of the patient’s clinical state. The reported differences for age require some caution in interpretation because of the narrow age range studied, but if confirmed in a broader population could provide a sensitive test of age-related vascular responsiveness. There is an intrinsic temptation to rush a new technique into different disease states without careful evaluation of influencing variables, a path which leads to overinterpretation and confusion among investigators. I commend the authors in their painstaking steps to define this technique more clearly in this study of healthy