Although H2(15)O is widely used for CBF measurement by positron tomography, it underestimates CBF, especially at elevated flow rates. Several tracers, including butanol, overcome this problem, but the short half-life of 15O provides advantages that cause water to remain the tracer of choice. We report the first use and evaluation of 15O-labeled butanol for CBF measurement. Flow measurements made in a similar fashion with water and butanol at 10-min intervals were compared in normal volunteers under resting and hypercapnic conditions. Regional analysis showed good agreement between the tracers at low flows, and significant underestimation of flow by water relative to butanol in regions of elevated flow. The observed relationship between the tracers and the curve-fitted permeability-surface area product for water (133 ml.100 g-1.min-1) follow the known relationship between water and true flow. These observations indicate that [15O]-butanol provided accurate measurements of human regional CBF under conditions of elevated perfusion. We conclude that butanol is a convenient and accurate method for routine CBF determination by positron emission tomography.