Objective To investigate the limitation of quantitative measurement of cerebral blood flow (CBF) of normal white matter by using a single subtraction with thin-slice TI1 periodic saturation (Q2TIPS Ⅱ ) pulsed arterial spin labeling (PASL)technique. Methods Thirty-one patients with brain tumors were examined at 3.0 T MRI system . A second version of quantitative imaging of perfusion using a single subtraction with additional thin-section periodic saturation after inversion and a time delay (Q2TIPS) technique of pulsed arterial spin labeling in the multisection mode and T2* dynamic susceptibility-weighted contrast-enhanced (T2* DSC)MR imaging were both implemented. Cerebral blood flow map obtained from PASL and DSC were reviewed. The regions of interest( ROI )were placed in the region of normal white matter contralateral to the lesion in the proximal and distal slices. In regions of interest, the signal intensity (SI)was measured from the maps of cerebral blood flow map obtained from PASL and DSC. Pair-t test was performed to determine if there were significant signal differences between proximal and distal slices. Pearson linear correlation analysis of signal intensity was performed for values from the same slices of PASL-CBF and DSC-CBF maps. Results In the deep white matter of distal slice, PASL-CBF map showed perfusion deficit while DSC-CBF map showed low CBF in the corresponding brain area. With the increased inversion time,the PASL-CBF map showed obviously improved perfusion signal in deep white matter (but still some perfusion deficit)and slightly decreased perfusion signal in grey matter. The mean signal of normal white matter measured from distal slices of PASL-CBF maps was( -22.1 ±55.5) ml· 100 g-1 · min-1 while it was (89.5 ±45.5) ml. 100 g-1 · min-1 in proximal slices. There was a significant difference of signal intensity from PASL-CBF maps between distal and proximal slices ( t = - 9. 512, P ＜ 0. 01 ＝, while no difference of signal intensity between distal[ (62. 8 ± 29.9) ml · 100 g-1 · min-1] and proximal slices [(57. 1 ±29.6) ml · 100 g-1 · min-1 ]was obtained from DSC-CBF maps(t= -1.607,P＞0.05). There was no significant correlation between PASL-CBF and DSC-CBF in both distal ( r = 0. 093, P ＞ 0. 05 ) and proximal slices ( r = - 0. 234, P ＞ 0. 05). ConclusionsPASL has limitation in the accurate quantification of cerebral blood flow of normal white matter. The quantification of CBF was affected by the limitations of the technique itself and the different parameters chosen..