Background: The goal of this study was to determine the reliability of TCD in evaluation of vertebrobasilar arteries in comparison with brain MRA in patients suffering from acute vertebrobasilar stroke. Methods: Samples were patients with definite clinical diagnosis of vertebrobasilar stroke. For all patients brain MRI, MRA and TCD were performed during the first 48 hours of admission. Basilar artery was insonated at the depth of 75 to 85 mm and vertebral arteries were insonated at the depth of 45 to 55 mm. On brain MRA, the degree of stenosis in vertebrobasilar arteries was graded from I (normal) to IV (total stenosis) and the correlation between the grade of stenosis and TCD indices were studied. Results: Spearman correlation test revealed a significant correlation between mean flow velocity (MFV) and MRA grading (correlation coefficient = -0.486) as well as end diastolic velocity (EDV) and MRA grading (correlation coefficient = -0.323) with no significant correlation between pulsatility index, peak systolic velocity and MRA grading (p > 0.05). One way ANOVA analysis showed that there was only significant mean MFV and mean EDV difference between grade 1 and other grades. Conclusions: TCD was only able to differentiate between stenotic and normal pattern and could not assist in the grading of stenosis. On the other hand, in acute vertebrobasilar stroke that TCD performed blindly without visualization of arteries and in a fixed depth it might have limited value in the grading of vertebrobasilar system stenosis.