Water capture mechanisms of zeolitic imidazolate framework ZIF-90 are revealed by differentiating the water clustering at interior interfaces of ZIF-90 and the center pore filling step, using vibrational sum-frequency generation spectroscopy (VSFG) at a one-micron spatial resolution. Spectral lineshapes of VSFG and IR spectra suggest that OD modes of heavy water in both water clustering and center pore filling steps experience similar environments, which is unexpected as weaker hydrogen bond interactions are involved in initial water clustering at interior surfaces. VSFG intensity shows similar dependence on the relative humidity as the adsorption isotherm, suggesting that water clustering and pore filling occur simultaneously. MD simulations based on MB-pol corroborate the experimental observations, indicating that water clustering and center pore filling happen nearly simultaneously within each pore, with water filling the other pores sequentially. The integration of nonlinear optics with computational simulations provides critical mechanistic insights into the pore filling mechanism that could be applied to the rational design of future MOFs.