The kinetics of caffeine uptake into the interlayer nanospace of silicate nanosheets modified with benzylammonium (BA) was evaluated by in situ monitoring the basal spacing in aqueous media using transmission X-ray diffraction. An interlayer spacing of 0.58 nm in water before caffeine uptake indicates a monomolecular layer of BA and a few water layers in each interlayer. The interlayer space expanded by 0.10 nm upon caffeine uptake (intercalation) and saturated even in the presence of excess caffeine. Time-course profiles of the interlayer spacing and the uptake amount after injection of caffeine into the water slurry were obtained. At the initial period, the plot for the basal spacing was located above that for the adsorbed amount, suggesting that the rate of the interlayer spacing change was faster than that to attain the adsorption equilibrium. A first-order kinetic simulation fitted to the profile also indicates that the basal spacing included a rapid expansion of 0.08 nm within a few minutes and a slow expansion of 0.02 nm over several hours. Regarding the slow component, the rate constant for the basal spacing was lower than that for the amount of caffeine adsorbed, meaning that a steady-state basal spacing is reached after the adsorption equilibrium.