Moisture sorption has a great impact on the mechanical properties of lignin. To better characterize the moisture sorption of lignin, an approach for in situ qualitative and quantitative analysis of moisture adsorption in nanogram-scaled lignin by using micro-FTIR spectroscopy and partial least squares regression is introduced in this study. Spectra of nanogram-scaled lignin were collected within the relative humidity (RH) of 0%-92%. A qualitative analysis of these measured spectra confirmed the effective water sorption sites and determined spectral ranges related to moisture adsorption. Using these identified spectral ranges, a quantitative forecasting model for the moisture content (MC) of lignin was built and developed according to partial least square regression (R2, 0.9996; RMSECV, 0.408; RMSEP, 0.118). Furthermore, the water sorption isotherm of lignin was acquired using the established forecasting model in which a very positive correlation between the estimated and measured MCs was demonstrated using a dynamic vapor sorption (DVS) setup. The results confirmed the practicability and effectiveness of this in situ qualitative and quantitative analysis approach.