Transcranial Doppler ultrasonography can monitor changes in intracranial blood flow velocity over time in a variety of experimental and clinical settings with excellent temporal resolution. Alterations in arterial carbon dioxide pressure exert a profound influence on blood flow velocity. Such changes exhibit important individual fluctuation depending on respiratory status. This limits the ability of transcranial Doppler to accurately study subtle changes in blood flow velocity, independent of the respiratory state of the subject. Suggested here is a method to control for the respiration artifact on blood flow velocity. The middle cerebral artery of 7 healthy male volunteers was studied with transcranial Doppler under resting conditions, monitoring end-tidal carbon dioxide concentration and blood flow velocity. Hyperventilation was performed both voluntarily and with pharmacological induction by human corticotropin-releasing hormone. These studies were carried out both with and without the use of counterregulation of the end-tidal carbon dioxide concentration via a respiration unit, with an adjustable carbon dioxide-oxygen gas supply preventing significant changes in end-tidal carbon dioxide. The blood flow velocity in the middle cerebral artery during maximal voluntary hyperventilation decreased from baseline values of 100% to 44.4 +/- 4.3% (a 55.6% decrease), and with human corticotropin-releasing hormone-induced involuntary hyperventilation, to 65.1 +/- 5.3% (a 34.9% decrease). With the control method, blood flow velocities during voluntary and pharmacological hyperventilation were 100 +/- 1.6% and 100 +/- 2.8%, respectively. This method allows for control of respiration-induced artifacts during transcranial Doppler monitoring, and can be used to assess the effect of direct or indirect blood flow velocity stimuli independent of respiratory status.