Functional MRI is based on the vascular response due to neuronal activation. The underlying mechanism of fMRI is the blood oxygenation level-dependent (BOLD) effect-a complex interplay between changes in the cerebral metabolisation rate of oxygen (CMRO2), neurovascular coupling, and the resulting hemodynamic response. An intact neurovascular coupling is essential for the detection of the BOLD signal and it seems likely that a disturbed cerebrovascular reserve capacity (CVRC) alters the BOLD response. We tested the hypothesis that extra- or intracranial artery disease influences the BOLD signal. Twenty-one patients with extra- or intracranial stenosis were studied with BOLD sensitive T2*-weighted MRI. All patients presented with transient or prolonged reversible ischemic symptoms ipsilateral to the artery disease but were asymptomatic at the time point of the MRI study. fMRI was performed employing a simple motor task (fist closure right and left). Additionally, the CVRC was assessed applying carbogen gas during serial T2*-weighted MRI for the calculation of CO(2) reactivity maps of the relative signal change. Signal differences between both hemispheres were compared in individual subjects and with healthy subjects. Patients with disturbed CVRC in the CO(2) reactivity maps showed either a significantly reduced (n = 5) or a negative (n = 1) BOLD signal in the affected compared to the unaffected primary sensorimotor cortex during fist closure. Patients with intact CVRC showed no significant BOLD signal differences between affected and unaffected hemisphere. Extra- or intracranial artery disease influences CVRC and consequently the BOLD signal. This observation is important for the clinical application of fMRI paradigms.