Pseudo-continuous arterial spin labeling (pCASL) with 3D non-segmented acquisition scheme allows assessment of full brain CBF with sufficiently high SNR and resolution in an efficient manner. Post-labeling delay is a critical parameter which needs to be determined preferably on a per subject basis since PLD can change based on the physiological condition of a subject. Multiple post-labeling delay (PLD) pCASL could address this issue. However, multiple PLD pCASL is not feasible for the 3D non-segmented acquisition scheme. In this work, we related the velocity in carotid artery (CA) measured using two schemes to PLD in ten volunteers. A relationship between the CA velocity and optimal PLD was derived which allows for optimal PLD to be determined on the fly for each subject.