The challenges in transferring and executing a near-infrared (NIR) spectroscopic method for croscarmellose (disintegrant) in binary blends for a continuous manufacturing (CM) process are presented. This work demonstrates the development of a NIR calibration model and its use to determine the blending parameters needed for binary blends at a development plant and later used to predict CM process blends. The calibration models were developed with laboratory scale powder blends ranging from 4.32%-64.77 (%w/w) of croscarmellose and evaluated using independent test blends. The selected model was then transferred to the continuous manufacturing development site to determine the croscarmellose concentration for spectra collected in real-time. A total of 18 development plant runs were monitored using an in-line NIR spectrometer, however, these spectra showed high baseline variations. The baseline variations were caused by the poor flow of the material within the system. An inconsistent bias which varied from 2.51 to 14.95 (%w/w) was observed in the predictions of croscarmellose. High baseline spectra were eliminated and the bias was significantly reduced by 42-51%. Experiments at lower flow rate speeds did not show significant changes in baseline and bias values showed more consistency. The calibration model was then transferred to two NIR spectrometers installed at-line at the commercial site, where powder samples were collected at the beginning middle and end of each CM plant run. The NIR calibration model predicted disintegrant concentration from the powder samples. Results showed the bias values for the NIR (1) varied from 0.74 to 2.21 (%w/w) and NIR (2) from 0.28 to 3.39 (%w/w). Average concentration values for both equipments were very close to the reference concentration values of 43.18 and 50.98 (%w/w). The study showed the model was able to identify flow issues, identified as baseline shifts, that could be used to alert to problems in the powder bed that may warrant diversion from a production line. These powder flow problems such as air gaps and inconsistent powder bed height affected the NIR spectra collected at the development plant and provided results with high bias. A lower bias was obtained in samples collected at line after blending.