Background The feasibility of the mapping of quantitative cerebral blood flow (CBF) named flow-sensitive altating inversion recovery exempting separate T1 measurement (FAIREST) is still controversial. This study aimed to evaluate the reliability of FAIREST in the measurement of regional CBF (rCBF) in healthy volunteers.Methods Eighteen healthy volunteers underwent magnetic resonance (MR) scanning with the sequence of FAIREST. While they were at rest, rCBF values were obtained in various brain regions of interest (ROIs). The same scheme was repeated on every subject after two weeks. Statistical analysis was made to determine the effect of location, scan and side on the measurement of rCBF.Results The mean CBF values were (122 ± 28) ml · (100 g)-1 · min-1 and (43 ±10) ml · (100 g)-1 · min-1 in the gray and white matter respectively. There was significant main effect of location (t=-12.5, P＜0.01), but no significant effect of side. Paired t-test of ROIs in the same slice showed no significant difference in most sites between two scans, except in the gray matter of the bilateral frontal lobes (t=2.18-2.34, P ＜0.05). However, the rCBF values of the same structure obtained from different slices showed a significant difference (t=-3.49,P＜0.01).Conclusion FAIREST is a reliable technique in the measurement of rCBF, but different imaging slice may affect the agreement of rCBF across the scans.