Carbon dioxide (CO2) and acetazolamide are increasingly being used as vasodilators to detect cerebrovascular reserve capacity in patients of chronic cerebrovascular disease. The functional cerebrovascular reserve or ability of cerebral vessels to lower their resistance in response to decrease in cerebral perfusion pressure is expressed as change in cerebral blood flow from baseline under a vasodilatory stimuli. Theoretically a vasodilator causing maximum vasodilation, and thereby expressing complete reserve capacity would be more suitable for such a purpose. We quantitatively compared the vasodilating effect of 5% CO2 inhalation and 1 g of intravenous acetazolamide by positron emission tomography. Cerebrovascular reserve was quantified in six patients with chronic cerebrovascular disease in the same sitting, using oxygen-15 labeled water (H2(15)O) positron emission tomography at rest, during 5% CO2 inhalation and after 1 g intravenous acetazolamide. A significant linear correlation in both nonlesion hemisphere (r = 0.701, p < 0.001) and in lesion hemisphere (r = 0.626, p < 0.005) was found between CO2 and acetazolamide for cerebrovascular reserve capacity. This correlation improved by considering cerebrovascular reserve per unit change in arterial carbon dioxide (r = 0.744, p < 0.001 in nonlesion hemisphere and r = 0.721, p < 0.001 in lesion hemisphere). The quantitative value of global reserve capacity was different by CO2 stimuli (5.2%) and acetazolamide (49.7%). Though a similar vasodilatory response is elicited by both vasodilators, acetazolamide seems to be more potent and therefore should be preferred to detect patients with exhausted cerebrovascular reserve capacity.