• In-line monitoring of fluidized bed foam granulations via NIR spectroscopy.
• Partial least squares regression model built to predict mean granule diameter.
• Model predictions used with batch statistical process control to track granule size.
Wet granulation-a unit operation involving mixing polymeric binders with powdered formulations-is well established in the pharmaceutical industry, playing a major role in the manufacturing of oral solid dosage forms and improving the physical properties of granules (size, density, shape factor, etc.) before tableting. The foaming properties of aqueous polymeric binders prove useful for binder delivery within the mixing vessel, with foamed binders leading to enhanced process efficiency (binder distribution, drying time, and temperature) and product quality (heat-sensitive components) during granulation. Given the importance of this stage in producing oral solid dosage forms, understanding the relationship between critical process parameters and critical quality attributes is essential. The process analytical technology (PAT) framework enables process design, analysis, and control and facilitates process development via in-line spectroscopy combined with multivariate data analysis to yield critical product information during the unit operation. Herein, we used in-line NIR spectroscopy to monitor granule size in foam granulations of a pharmaceutical compound. The mean granule diameter was predicted using a partial least squares regression (PLSR) model (with a prediction error of 11.8 μm) and combined with a batch statistical process control (BSPC) approach for the temporal monitoring of granule size during three foam granulations.