A new method has been developed for diffusible tracers, to quantify CBF at rest and after pharmacological stress from a single session of dynamic scans with dual bolus administration of a radiotracer. The calculation process consisted of three steps, including the procedures of incorporating background radioactivity contaminated from the previous scan. Feasibility of this approach was tested on clinical SPECT studies on 16 subjects. Two sequential SPECT scans, 30 min apart, were carried out on each subject, after each of two split-dose administrations of 111 MBq IMP. Of these, 11 subjects received acetazolamide at 10 min before the second IMP injection. Additional PET scans were also carried out on 6 subjects on a separate day, at rest and after acetazolamide administration. The other 5 subjects were scanned only at rest during the whole study period. Quantitative CBF obtained by this method was in a good agreement with those determined with PET (y(ml/100 g/min)=1.07x(ml/100 g/min)-1.14, r=0.94). Vasareactivity was approximately 40% over the whole cerebral area on healthy controls, which was consistent with a literature value. Reproducibility of CBF determined in the rest-rest study was 1.5+/-5.7%. Noise enhancement of CBF images, particularly the second CBF, was reduced, providing reasonable image quality. Repeat assessment of quantitative CBF from a single session of scans with split-dose IMP is accurate, and may be applied to clinical research for assessing vascular reactivity in patients with chronic cerebral vascular disease.