We compared pharmacologically-perturbed hemodynamic parameters (cerebral blood volume; CBV, and flow; CBF) by acetazolamide administration in six healthy human subjects studied with positron emission tomography (PET) and near-infrared (NIR) time-resolved spectroscopy (TRS) simultaneously to investigate whether NIR-TRS could measure in vivo hemodynamics in the brain tissue quantitatively. Simultaneously with the PET measurements, TRS measurements were performed at the forehead with four different optode spacing from 2 cm to 5 cm. Total hemoglobin and oxygen saturation (SO2) measured by TRS significantly increased after administration of acetazolamide at any optode spacing in all subjects. In PET study, CBV and CBF were estimated in the following three volumes of interest (VOIs) determined on magnetic resonance images, VOI1: scalp and skull, VOI2: gray matter region, VOI3: gray and white matter regions. Acetazolamide treatment elevated CBF and CBV significantly in VOI2 and VOI3 but VOI1. TRS-derived CBV was more strongly correlated with PET-derived counterpart in VOI2 and VOI3 when the optode spacing was above 4 cm, although optical signal from cerebral tissue could be caught with any optode spacing. As to increase of the CBV, 4 cm of optode spacing correlated best with VOI2. To support the result of TRS-PET experiment, we also estimated the contribution ratios of intracerebral tissue to observed absorption change based on diffusion theory. The contribution ratios at 4 cm were estimated as follows: 761 nm: 50%, 791 nm: 72%, 836 nm: 70%. These results demonstrated that NIR-TRS with 4 cm of optode spacing could measure cerebral hemodynamic responses optimally and quantitatively.