The circle of Willis plays an important role in the distribution of blood flow in the brain. To obtain dynamic information of the blood flow through the circle of Willis, a dynamic MRA technique based on arterial spin labeling (ASL) is introduced as a non-invasive technique. When the ASL labeling slab is restricted to a single artery, it is possible to visualize selectively the flow distribution of that specific artery. However, because of the decay of the label and the presence of noise it is difficult to extract functional information from these images. In the present study we propose three visualization and post-processing methods for the interpretation of these images. Firstly, the passage of labeled blood was corrected for decay of the label and hereafter shown as a movie. Secondly, by calculating the time of arrival at every location in the arteries of the circle of Willis, a 2D image was reconstructed summarizing the information of the movie. Finally, quantitative flow values were obtained by relating the arterial input function to the passage of labeled blood through a region of interest encompassing the vessel under investigation. Experiments in a circle of Willis phantom showed a high linear relation between measured flow and true flow, although the measured values were 10-15% lower than the true flow values. Measurements in healthy volunteers showed the potential to quantify the flow in all major arteries of the circle of Willis.