The effects of ethanol on acoustically stimulated blood oxygenation level-dependent (BOLD) signal response in healthy humans was examined with echo planar functional magnetic resonance imaging (fMRI). An acquisition mode minimizing neuronal activation by scanner noise in combination with acoustic excitation by a pulsed 1000-Hz sine tone was used. Paradigms were repeated three times before and after the ingestion of 0.7 g of ethanol/kg(body weight). Linear correlation analyses (r>/=0.40) revealed bilateral BOLD responses in the auditory cortex. Significant voxels covered a cortical volume of approximately 3 ml that was reduced by approximately 40% after ethanol. The BOLD signal change initially reaching approximately 3% was reduced by 12-27%, depending on the definition of the region of interest for signal quantitation. Because ethanol produces vasodilation, the hemodynamic contribution to the BOLD signal change was estimated by modeling the relationship between regional cerebral blood flow (rCBF) and BOLD signal changes. Assuming a baseline flow increase by 10% after ethanol intake, the resulting 'Flow-BOLD-Dependence' (FBD) curve suggested that the ethanol-related BOLD signal reduction was approximately 7-12% greater than the reduction contributed purely by vasodilation. However, simultaneous determination of rCBF and regional cerebral blood volume would be required for an exact quantitation of the neuronally induced BOLD response. Although the FBD model needs empirical validation, its cautious implementation appears to be helpful if fMRI is used in combination with vasoactive drugs.